1 acpi= [HW,ACPI,X86,ARM64]
2 Advanced Configuration and Power Interface
3 Format: { force | on | off | strict | noirq | rsdt |
5 force -- enable ACPI if default was off
6 on -- enable ACPI but allow fallback to DT [arm64]
7 off -- disable ACPI if default was on
8 noirq -- do not use ACPI for IRQ routing
9 strict -- Be less tolerant of platforms that are not
10 strictly ACPI specification compliant.
11 rsdt -- prefer RSDT over (default) XSDT
12 copy_dsdt -- copy DSDT to memory
13 For ARM64, ONLY "acpi=off", "acpi=on" or "acpi=force"
16 See also Documentation/power/runtime_pm.rst, pci=noacpi
18 acpi_apic_instance= [ACPI, IOAPIC]
20 2: use 2nd APIC table, if available
21 1,0: use 1st APIC table
24 acpi_backlight= [HW,ACPI]
25 { vendor | video | native | none }
26 If set to vendor, prefer vendor-specific driver
27 (e.g. thinkpad_acpi, sony_acpi, etc.) instead
28 of the ACPI video.ko driver.
29 If set to video, use the ACPI video.ko driver.
30 If set to native, use the device's native backlight mode.
31 If set to none, disable the ACPI backlight interface.
33 acpi_force_32bit_fadt_addr
34 force FADT to use 32 bit addresses rather than the
35 64 bit X_* addresses. Some firmware have broken 64
36 bit addresses for force ACPI ignore these and use
37 the older legacy 32 bit addresses.
39 acpica_no_return_repair [HW, ACPI]
40 Disable AML predefined validation mechanism
41 This mechanism can repair the evaluation result to make
42 the return objects more ACPI specification compliant.
43 This option is useful for developers to identify the
44 root cause of an AML interpreter issue when the issue
45 has something to do with the repair mechanism.
47 acpi.debug_layer= [HW,ACPI,ACPI_DEBUG]
48 acpi.debug_level= [HW,ACPI,ACPI_DEBUG]
50 CONFIG_ACPI_DEBUG must be enabled to produce any ACPI
51 debug output. Bits in debug_layer correspond to a
52 _COMPONENT in an ACPI source file, e.g.,
53 #define _COMPONENT ACPI_PCI_COMPONENT
54 Bits in debug_level correspond to a level in
55 ACPI_DEBUG_PRINT statements, e.g.,
56 ACPI_DEBUG_PRINT((ACPI_DB_INFO, ...
57 The debug_level mask defaults to "info". See
58 Documentation/firmware-guide/acpi/debug.rst for more information about
59 debug layers and levels.
61 Enable processor driver info messages:
62 acpi.debug_layer=0x20000000
63 Enable PCI/PCI interrupt routing info messages:
64 acpi.debug_layer=0x400000
65 Enable AML "Debug" output, i.e., stores to the Debug
66 object while interpreting AML:
67 acpi.debug_layer=0xffffffff acpi.debug_level=0x2
68 Enable all messages related to ACPI hardware:
69 acpi.debug_layer=0x2 acpi.debug_level=0xffffffff
71 Some values produce so much output that the system is
72 unusable. The "log_buf_len" parameter may be useful
73 if you need to capture more output.
75 acpi_enforce_resources= [ACPI]
77 Check for resource conflicts between native drivers
78 and ACPI OperationRegions (SystemIO and SystemMemory
79 only). IO ports and memory declared in ACPI might be
80 used by the ACPI subsystem in arbitrary AML code and
81 can interfere with legacy drivers.
82 strict (default): access to resources claimed by ACPI
83 is denied; legacy drivers trying to access reserved
84 resources will fail to bind to device using them.
85 lax: access to resources claimed by ACPI is allowed;
86 legacy drivers trying to access reserved resources
87 will bind successfully but a warning message is logged.
88 no: ACPI OperationRegions are not marked as reserved,
89 no further checks are performed.
91 acpi_force_table_verification [HW,ACPI]
92 Enable table checksum verification during early stage.
93 By default, this is disabled due to x86 early mapping
96 acpi_irq_balance [HW,ACPI]
97 ACPI will balance active IRQs
100 acpi_irq_nobalance [HW,ACPI]
101 ACPI will not move active IRQs (default)
104 acpi_irq_isa= [HW,ACPI] If irq_balance, mark listed IRQs used by ISA
105 Format: <irq>,<irq>...
107 acpi_irq_pci= [HW,ACPI] If irq_balance, clear listed IRQs for
109 Format: <irq>,<irq>...
111 acpi_mask_gpe= [HW,ACPI]
112 Due to the existence of _Lxx/_Exx, some GPEs triggered
113 by unsupported hardware/firmware features can result in
114 GPE floodings that cannot be automatically disabled by
116 This facility can be used to prevent such uncontrolled
120 acpi_no_auto_serialize [HW,ACPI]
121 Disable auto-serialization of AML methods
122 AML control methods that contain the opcodes to create
123 named objects will be marked as "Serialized" by the
124 auto-serialization feature.
125 This feature is enabled by default.
126 This option allows to turn off the feature.
128 acpi_no_memhotplug [ACPI] Disable memory hotplug. Useful for kdump
131 acpi_no_static_ssdt [HW,ACPI]
132 Disable installation of static SSDTs at early boot time
133 By default, SSDTs contained in the RSDT/XSDT will be
134 installed automatically and they will appear under
135 /sys/firmware/acpi/tables.
136 This option turns off this feature.
137 Note that specifying this option does not affect
138 dynamic table installation which will install SSDT
139 tables to /sys/firmware/acpi/tables/dynamic.
141 acpi_no_watchdog [HW,ACPI,WDT]
142 Ignore the ACPI-based watchdog interface (WDAT) and let
143 a native driver control the watchdog device instead.
145 acpi_rsdp= [ACPI,EFI,KEXEC]
146 Pass the RSDP address to the kernel, mostly used
147 on machines running EFI runtime service to boot the
148 second kernel for kdump.
150 acpi_os_name= [HW,ACPI] Tell ACPI BIOS the name of the OS
151 Format: To spoof as Windows 98: ="Microsoft Windows"
153 acpi_rev_override [ACPI] Override the _REV object to return 5 (instead
154 of 2 which is mandated by ACPI 6) as the supported ACPI
155 specification revision (when using this switch, it may
156 be necessary to carry out a cold reboot _twice_ in a
157 row to make it take effect on the platform firmware).
159 acpi_osi= [HW,ACPI] Modify list of supported OS interface strings
160 acpi_osi="string1" # add string1
161 acpi_osi="!string2" # remove string2
162 acpi_osi=!* # remove all strings
163 acpi_osi=! # disable all built-in OS vendor
165 acpi_osi=!! # enable all built-in OS vendor
167 acpi_osi= # disable all strings
169 'acpi_osi=!' can be used in combination with single or
170 multiple 'acpi_osi="string1"' to support specific OS
171 vendor string(s). Note that such command can only
172 affect the default state of the OS vendor strings, thus
173 it cannot affect the default state of the feature group
174 strings and the current state of the OS vendor strings,
175 specifying it multiple times through kernel command line
176 is meaningless. This command is useful when one do not
177 care about the state of the feature group strings which
178 should be controlled by the OSPM.
180 1. 'acpi_osi=! acpi_osi="Windows 2000"' is equivalent
181 to 'acpi_osi="Windows 2000" acpi_osi=!', they all
182 can make '_OSI("Windows 2000")' TRUE.
184 'acpi_osi=' cannot be used in combination with other
185 'acpi_osi=' command lines, the _OSI method will not
186 exist in the ACPI namespace. NOTE that such command can
187 only affect the _OSI support state, thus specifying it
188 multiple times through kernel command line is also
191 1. 'acpi_osi=' can make 'CondRefOf(_OSI, Local1)'
194 'acpi_osi=!*' can be used in combination with single or
195 multiple 'acpi_osi="string1"' to support specific
196 string(s). Note that such command can affect the
197 current state of both the OS vendor strings and the
198 feature group strings, thus specifying it multiple times
199 through kernel command line is meaningful. But it may
200 still not able to affect the final state of a string if
201 there are quirks related to this string. This command
202 is useful when one want to control the state of the
203 feature group strings to debug BIOS issues related to
206 1. 'acpi_osi="Module Device" acpi_osi=!*' can make
207 '_OSI("Module Device")' FALSE.
208 2. 'acpi_osi=!* acpi_osi="Module Device"' can make
209 '_OSI("Module Device")' TRUE.
210 3. 'acpi_osi=! acpi_osi=!* acpi_osi="Windows 2000"' is
212 'acpi_osi=!* acpi_osi=! acpi_osi="Windows 2000"'
214 'acpi_osi=!* acpi_osi="Windows 2000" acpi_osi=!',
215 they all will make '_OSI("Windows 2000")' TRUE.
218 Override the pmtimer bug detection: force the kernel
219 to assume that this machine's pmtimer latches its value
220 and always returns good values.
222 acpi_sci= [HW,ACPI] ACPI System Control Interrupt trigger mode
223 Format: { level | edge | high | low }
225 acpi_skip_timer_override [HW,ACPI]
226 Recognize and ignore IRQ0/pin2 Interrupt Override.
227 For broken nForce2 BIOS resulting in XT-PIC timer.
229 acpi_sleep= [HW,ACPI] Sleep options
230 Format: { s3_bios, s3_mode, s3_beep, s4_nohwsig,
231 old_ordering, nonvs, sci_force_enable, nobl }
232 See Documentation/power/video.rst for information on
234 s3_beep is for debugging; it makes the PC's speaker beep
235 as soon as the kernel's real-mode entry point is called.
236 s4_nohwsig prevents ACPI hardware signature from being
237 used during resume from hibernation.
238 old_ordering causes the ACPI 1.0 ordering of the _PTS
239 control method, with respect to putting devices into
240 low power states, to be enforced (the ACPI 2.0 ordering
241 of _PTS is used by default).
242 nonvs prevents the kernel from saving/restoring the
243 ACPI NVS memory during suspend/hibernation and resume.
244 sci_force_enable causes the kernel to set SCI_EN directly
245 on resume from S1/S3 (which is against the ACPI spec,
246 but some broken systems don't work without it).
247 nobl causes the internal blacklist of systems known to
248 behave incorrectly in some ways with respect to system
249 suspend and resume to be ignored (use wisely).
251 acpi_use_timer_override [HW,ACPI]
252 Use timer override. For some broken Nvidia NF5 boards
253 that require a timer override, but don't have HPET
255 add_efi_memmap [EFI; X86] Include EFI memory map in
256 kernel's map of available physical RAM.
259 { off | try_unsupported }
260 off: disable AGP support
261 try_unsupported: try to drive unsupported chipsets
262 (may crash computer or cause data corruption)
265 See Documentation/sound/alsa-configuration.rst
268 Allow the default userspace alignment fault handler
269 behaviour to be specified. Bit 0 enables warnings,
270 bit 1 enables fixups, and bit 2 sends a segfault.
272 align_va_addr= [X86-64]
273 Align virtual addresses by clearing slice [14:12] when
274 allocating a VMA at process creation time. This option
275 gives you up to 3% performance improvement on AMD F15h
276 machines (where it is enabled by default) for a
277 CPU-intensive style benchmark, and it can vary highly in
278 a microbenchmark depending on workload and compiler.
280 32: only for 32-bit processes
281 64: only for 64-bit processes
282 on: enable for both 32- and 64-bit processes
283 off: disable for both 32- and 64-bit processes
285 alloc_snapshot [FTRACE]
286 Allocate the ftrace snapshot buffer on boot up when the
287 main buffer is allocated. This is handy if debugging
288 and you need to use tracing_snapshot() on boot up, and
289 do not want to use tracing_snapshot_alloc() as it needs
290 to be done where GFP_KERNEL allocations are allowed.
292 amd_iommu= [HW,X86-64]
293 Pass parameters to the AMD IOMMU driver in the system.
295 fullflush - enable flushing of IO/TLB entries when
296 they are unmapped. Otherwise they are
297 flushed before they will be reused, which
299 off - do not initialize any AMD IOMMU found in
301 force_isolation - Force device isolation for all
302 devices. The IOMMU driver is not
303 allowed anymore to lift isolation
304 requirements as needed. This option
305 does not override iommu=pt
307 amd_iommu_dump= [HW,X86-64]
308 Enable AMD IOMMU driver option to dump the ACPI table
309 for AMD IOMMU. With this option enabled, AMD IOMMU
310 driver will print ACPI tables for AMD IOMMU during
311 IOMMU initialization.
313 amd_iommu_intr= [HW,X86-64]
314 Specifies one of the following AMD IOMMU interrupt
316 legacy - Use legacy interrupt remapping mode.
317 vapic - Use virtual APIC mode, which allows IOMMU
318 to inject interrupts directly into guest.
319 This mode requires kvm-amd.avic=1.
320 (Default when IOMMU HW support is present.)
322 amijoy.map= [HW,JOY] Amiga joystick support
323 Map of devices attached to JOY0DAT and JOY1DAT
325 See also Documentation/input/joydev/joystick.rst
327 analog.map= [HW,JOY] Analog joystick and gamepad support
328 Specifies type or capabilities of an analog joystick
329 connected to one of 16 gameports
330 Format: <type1>,<type2>,..<type16>
333 Power management functions (SPARCstation-4/5 + deriv.)
335 Disable APC CPU standby support. SPARCstation-Fox does
336 not play well with APC CPU idle - disable it if you have
337 APC and your system crashes randomly.
339 apic= [APIC,X86] Advanced Programmable Interrupt Controller
340 Change the output verbosity while booting
341 Format: { quiet (default) | verbose | debug }
342 Change the amount of debugging information output
343 when initialising the APIC and IO-APIC components.
344 For X86-32, this can also be used to specify an APIC
346 Format: apic=driver_name
347 Examples: apic=bigsmp
349 apic_extnmi= [APIC,X86] External NMI delivery setting
350 Format: { bsp (default) | all | none }
351 bsp: External NMI is delivered only to CPU 0
352 all: External NMIs are broadcast to all CPUs as a
354 none: External NMI is masked for all CPUs. This is
355 useful so that a dump capture kernel won't be
359 See Documentation/networking/ipv6.txt.
361 show_lapic= [APIC,X86] Advanced Programmable Interrupt Controller
362 Limit apic dumping. The parameter defines the maximal
363 number of local apics being dumped. Also it is possible
364 to set it to "all" by meaning -- no limit here.
365 Format: { 1 (default) | 2 | ... | all }.
366 The parameter valid if only apic=debug or
367 apic=verbose is specified.
368 Example: apic=debug show_lapic=all
370 apm= [APM] Advanced Power Management
371 See header of arch/x86/kernel/apm_32.c.
373 arcrimi= [HW,NET] ARCnet - "RIM I" (entirely mem-mapped) cards
374 Format: <io>,<irq>,<nodeID>
378 atarimouse= [HW,MOUSE] Atari Mouse
380 atkbd.extra= [HW] Enable extra LEDs and keys on IBM RapidAccess,
381 EzKey and similar keyboards
383 atkbd.reset= [HW] Reset keyboard during initialization
385 atkbd.set= [HW] Select keyboard code set
386 Format: <int> (2 = AT (default), 3 = PS/2)
388 atkbd.scroll= [HW] Enable scroll wheel on MS Office and similar
391 atkbd.softraw= [HW] Choose between synthetic and real raw mode
392 Format: <bool> (0 = real, 1 = synthetic (default))
394 atkbd.softrepeat= [HW]
395 Use software keyboard repeat
397 audit= [KNL] Enable the audit sub-system
398 Format: { "0" | "1" | "off" | "on" }
399 0 | off - kernel audit is disabled and can not be
400 enabled until the next reboot
401 unset - kernel audit is initialized but disabled and
402 will be fully enabled by the userspace auditd.
403 1 | on - kernel audit is initialized and partially
404 enabled, storing at most audit_backlog_limit
405 messages in RAM until it is fully enabled by the
409 audit_backlog_limit= [KNL] Set the audit queue size limit.
410 Format: <int> (must be >=0)
413 bau= [X86_UV] Enable the BAU on SGI UV. The default
414 behavior is to disable the BAU (i.e. bau=0).
415 Format: { "0" | "1" }
418 unset - Disable the BAU.
420 baycom_epp= [HW,AX25]
423 baycom_par= [HW,AX25] BayCom Parallel Port AX.25 Modem
425 See header of drivers/net/hamradio/baycom_par.c.
427 baycom_ser_fdx= [HW,AX25]
428 BayCom Serial Port AX.25 Modem (Full Duplex Mode)
429 Format: <io>,<irq>,<mode>[,<baud>]
430 See header of drivers/net/hamradio/baycom_ser_fdx.c.
432 baycom_ser_hdx= [HW,AX25]
433 BayCom Serial Port AX.25 Modem (Half Duplex Mode)
434 Format: <io>,<irq>,<mode>
435 See header of drivers/net/hamradio/baycom_ser_hdx.c.
437 blkdevparts= Manual partition parsing of block device(s) for
438 embedded devices based on command line input.
439 See Documentation/block/cmdline-partition.rst
441 boot_delay= Milliseconds to delay each printk during boot.
442 Values larger than 10 seconds (10000) are changed to
447 Extended command line options can be added to an initrd
448 and this will cause the kernel to look for it.
450 See Documentation/admin-guide/bootconfig.rst
453 Disable BERT OS support on buggy BIOSes.
455 bgrt_disable [ACPI][X86]
456 Disable BGRT to avoid flickering OEM logo.
458 bttv.card= [HW,V4L] bttv (bt848 + bt878 based grabber cards)
459 bttv.radio= Most important insmod options are available as
461 bttv.pll= See Documentation/media/v4l-drivers/bttv.rst
464 bulk_remove=off [PPC] This parameter disables the use of the pSeries
465 firmware feature for flushing multiple hpte entries
468 c101= [NET] Moxa C101 synchronous serial card
470 cachesize= [BUGS=X86-32] Override level 2 CPU cache size detection.
471 Sometimes CPU hardware bugs make them report the cache
472 size incorrectly. The kernel will attempt work arounds
473 to fix known problems, but for some CPUs it is not
474 possible to determine what the correct size should be.
475 This option provides an override for these situations.
478 [NET] Specifies amount of time (in seconds) that
479 the kernel should wait for a network carrier. By default
480 it waits 120 seconds.
482 ca_keys= [KEYS] This parameter identifies a specific key(s) on
483 the system trusted keyring to be used for certificate
485 format: { id:<keyid> | builtin }
487 cca= [MIPS] Override the kernel pages' cache coherency
488 algorithm. Accepted values range from 0 to 7
489 inclusive. See arch/mips/include/asm/pgtable-bits.h
490 for platform specific values (SB1, Loongson3 and
493 ccw_timeout_log [S390]
494 See Documentation/s390/common_io.rst for details.
496 cgroup_disable= [KNL] Disable a particular controller
497 Format: {name of the controller(s) to disable}
498 The effects of cgroup_disable=foo are:
499 - foo isn't auto-mounted if you mount all cgroups in
501 - foo isn't visible as an individually mountable
503 {Currently only "memory" controller deal with this and
504 cut the overhead, others just disable the usage. So
505 only cgroup_disable=memory is actually worthy}
507 cgroup_no_v1= [KNL] Disable cgroup controllers and named hierarchies in v1
508 Format: { { controller | "all" | "named" }
509 [,{ controller | "all" | "named" }...] }
510 Like cgroup_disable, but only applies to cgroup v1;
511 the blacklisted controllers remain available in cgroup2.
512 "all" blacklists all controllers and "named" disables
513 named mounts. Specifying both "all" and "named" disables
516 cgroup.memory= [KNL] Pass options to the cgroup memory controller.
518 nosocket -- Disable socket memory accounting.
519 nokmem -- Disable kernel memory accounting.
521 checkreqprot [SELINUX] Set initial checkreqprot flag value.
522 Format: { "0" | "1" }
523 See security/selinux/Kconfig help text.
524 0 -- check protection applied by kernel (includes
525 any implied execute protection).
526 1 -- check protection requested by application.
527 Default value is set via a kernel config option.
528 Value can be changed at runtime via
529 /sys/fs/selinux/checkreqprot.
530 Setting checkreqprot to 1 is deprecated.
533 See Documentation/s390/common_io.rst for details.
536 Prevents the clock framework from automatically gating
537 clocks that have not been explicitly enabled by a Linux
538 device driver but are enabled in hardware at reset or
539 by the bootloader/firmware. Note that this does not
540 force such clocks to be always-on nor does it reserve
541 those clocks in any way. This parameter is useful for
542 debug and development, but should not be needed on a
543 platform with proper driver support. For more
544 information, see Documentation/driver-api/clk.rst.
546 clock= [BUGS=X86-32, HW] gettimeofday clocksource override.
548 Forces specified clocksource (if available) to be used
549 when calculating gettimeofday(). If specified
550 clocksource is not available, it defaults to PIT.
551 Format: { pit | tsc | cyclone | pmtmr }
553 clocksource= Override the default clocksource
555 Override the default clocksource and use the clocksource
556 with the name specified.
557 Some clocksource names to choose from, depending on
559 [all] jiffies (this is the base, fallback clocksource)
561 [ARM] imx_timer1,OSTS,netx_timer,mpu_timer2,
562 pxa_timer,timer3,32k_counter,timer0_1
563 [X86-32] pit,hpet,tsc;
564 scx200_hrt on Geode; cyclone on IBM x440
572 clocksource.arm_arch_timer.evtstrm=
575 Enable/disable the eventstream feature of the ARM
576 architected timer so that code using WFE-based polling
577 loops can be debugged more effectively on production
580 clearcpuid=BITNUM [X86]
581 Disable CPUID feature X for the kernel. See
582 arch/x86/include/asm/cpufeatures.h for the valid bit
583 numbers. Note the Linux specific bits are not necessarily
584 stable over kernel options, but the vendor specific
586 Also note that user programs calling CPUID directly
587 or using the feature without checking anything
588 will still see it. This just prevents it from
589 being used by the kernel or shown in /proc/cpuinfo.
590 Also note the kernel might malfunction if you disable
593 cma=nn[MG]@[start[MG][-end[MG]]]
595 Sets the size of kernel global memory area for
596 contiguous memory allocations and optionally the
597 placement constraint by the physical address range of
598 memory allocations. A value of 0 disables CMA
599 altogether. For more information, see
600 include/linux/dma-contiguous.h
602 cmo_free_hint= [PPC] Format: { yes | no }
603 Specify whether pages are marked as being inactive
604 when they are freed. This is used in CMO environments
605 to determine OS memory pressure for page stealing by
609 coherent_pool=nn[KMG] [ARM,KNL]
610 Sets the size of memory pool for coherent, atomic dma
611 allocations, by default set to 256K.
613 com20020= [HW,NET] ARCnet - COM20020 chipset
615 <io>[,<irq>[,<nodeID>[,<backplane>[,<ckp>[,<timeout>]]]]]
617 com90io= [HW,NET] ARCnet - COM90xx chipset (IO-mapped buffers)
621 ARCnet - COM90xx chipset (memory-mapped buffers)
622 Format: <io>[,<irq>[,<memstart>]]
624 condev= [HW,S390] console device
627 console= [KNL] Output console device and options.
629 tty<n> Use the virtual console device <n>.
633 Use the specified serial port. The options are of
634 the form "bbbbpnf", where "bbbb" is the baud rate,
635 "p" is parity ("n", "o", or "e"), "n" is number of
636 bits, and "f" is flow control ("r" for RTS or
637 omit it). Default is "9600n8".
639 See Documentation/admin-guide/serial-console.rst for more
641 Documentation/networking/netconsole.txt for an
644 uart[8250],io,<addr>[,options]
645 uart[8250],mmio,<addr>[,options]
646 uart[8250],mmio16,<addr>[,options]
647 uart[8250],mmio32,<addr>[,options]
648 uart[8250],0x<addr>[,options]
649 Start an early, polled-mode console on the 8250/16550
650 UART at the specified I/O port or MMIO address,
651 switching to the matching ttyS device later.
652 MMIO inter-register address stride is either 8-bit
653 (mmio), 16-bit (mmio16), or 32-bit (mmio32).
654 If none of [io|mmio|mmio16|mmio32], <addr> is assumed
655 to be equivalent to 'mmio'. 'options' are specified in
656 the same format described for ttyS above; if unspecified,
657 the h/w is not re-initialized.
659 hvc<n> Use the hypervisor console device <n>. This is for
660 both Xen and PowerPC hypervisors.
662 If the device connected to the port is not a TTY but a braille
663 device, prepend "brl," before the device type, for instance
665 For now, only VisioBraille is supported.
668 [KNL] Change console messages format
670 By default we print messages on consoles in
671 "[time stamp] text\n" format (time stamp may not be
672 printed, depending on CONFIG_PRINTK_TIME or
673 `printk_time' param).
675 Switch to syslog format: "<%u>[time stamp] text\n"
676 IOW, each message will have a facility and loglevel
677 prefix. The format is similar to one used by syslog()
678 syscall, or to executing "dmesg -S --raw" or to reading
681 consoleblank= [KNL] The console blank (screen saver) timeout in
682 seconds. A value of 0 disables the blank timer.
686 [KNL] Change the default value for
687 /proc/<pid>/coredump_filter.
688 See also Documentation/filesystems/proc.rst.
690 coresight_cpu_debug.enable
693 Enable/disable the CPU sampling based debugging.
694 0: default value, disable debugging
695 1: enable debugging at boot time
697 cpuidle.off=1 [CPU_IDLE]
698 disable the cpuidle sub-system
701 [CPU_IDLE] Name of the cpuidle governor to use.
703 cpufreq.off=1 [CPU_FREQ]
704 disable the cpufreq sub-system
707 [X86] Delay for N microsec between assert and de-assert
708 of APIC INIT to start processors. This delay occurs
709 on every CPU online, such as boot, and resume from suspend.
712 cpcihp_generic= [HW,PCI] Generic port I/O CompactPCI driver
714 <first_slot>,<last_slot>,<port>,<enum_bit>[,<debug>]
716 crashkernel=size[KMG][@offset[KMG]]
717 [KNL] Using kexec, Linux can switch to a 'crash kernel'
718 upon panic. This parameter reserves the physical
719 memory region [offset, offset + size] for that kernel
720 image. If '@offset' is omitted, then a suitable offset
721 is selected automatically.
722 [KNL, x86_64] select a region under 4G first, and
723 fall back to reserve region above 4G when '@offset'
724 hasn't been specified.
725 See Documentation/admin-guide/kdump/kdump.rst for further details.
727 crashkernel=range1:size1[,range2:size2,...][@offset]
728 [KNL] Same as above, but depends on the memory
729 in the running system. The syntax of range is
730 start-[end] where start and end are both
731 a memory unit (amount[KMG]). See also
732 Documentation/admin-guide/kdump/kdump.rst for an example.
734 crashkernel=size[KMG],high
735 [KNL, x86_64] range could be above 4G. Allow kernel
736 to allocate physical memory region from top, so could
737 be above 4G if system have more than 4G ram installed.
738 Otherwise memory region will be allocated below 4G, if
740 It will be ignored if crashkernel=X is specified.
741 crashkernel=size[KMG],low
742 [KNL, x86_64] range under 4G. When crashkernel=X,high
743 is passed, kernel could allocate physical memory region
744 above 4G, that cause second kernel crash on system
745 that require some amount of low memory, e.g. swiotlb
746 requires at least 64M+32K low memory, also enough extra
747 low memory is needed to make sure DMA buffers for 32-bit
748 devices won't run out. Kernel would try to allocate at
749 at least 256M below 4G automatically.
750 This one let user to specify own low range under 4G
751 for second kernel instead.
752 0: to disable low allocation.
753 It will be ignored when crashkernel=X,high is not used
754 or memory reserved is below 4G.
757 [KNL] Disable crypto self-tests
762 cs89x0_media= [HW,NET]
763 Format: { rj45 | aui | bnc }
766 See header of drivers/s390/block/dasd_devmap.c.
768 db9.dev[2|3]= [HW,JOY] Multisystem joystick support via parallel port
769 (one device per port)
770 Format: <port#>,<type>
771 See also Documentation/input/devices/joystick-parport.rst
773 ddebug_query= [KNL,DYNAMIC_DEBUG] Enable debug messages at early boot
775 Documentation/admin-guide/dynamic-debug-howto.rst for
776 details. Deprecated, see dyndbg.
778 debug [KNL] Enable kernel debugging (events log level).
781 [KNL] Enable printing [hashed] pointers early in the
782 boot sequence. If enabled, we use a weak hash instead
783 of siphash to hash pointers. Use this option if you are
784 seeing instances of '(___ptrval___)') and need to see a
785 value (hashed pointer) instead. Cryptographically
786 insecure, please do not use on production kernels.
789 [KNL] verbose self-tests
791 Print debugging info while doing the locking API
793 We default to 0 (no extra messages), setting it to
794 1 will print _a lot_ more information - normally
795 only useful to kernel developers.
797 debug_objects [KNL] Enable object debugging
800 [KNL] Disable object debugging
802 debug_guardpage_minorder=
803 [KNL] When CONFIG_DEBUG_PAGEALLOC is set, this
804 parameter allows control of the order of pages that will
805 be intentionally kept free (and hence protected) by the
806 buddy allocator. Bigger value increase the probability
807 of catching random memory corruption, but reduce the
808 amount of memory for normal system use. The maximum
809 possible value is MAX_ORDER/2. Setting this parameter
810 to 1 or 2 should be enough to identify most random
811 memory corruption problems caused by bugs in kernel or
812 driver code when a CPU writes to (or reads from) a
813 random memory location. Note that there exists a class
814 of memory corruptions problems caused by buggy H/W or
815 F/W or by drivers badly programing DMA (basically when
816 memory is written at bus level and the CPU MMU is
817 bypassed) which are not detectable by
818 CONFIG_DEBUG_PAGEALLOC, hence this option will not help
819 tracking down these problems.
822 [KNL] When CONFIG_DEBUG_PAGEALLOC is set, this parameter
823 enables the feature at boot time. By default, it is
824 disabled and the system will work mostly the same as a
825 kernel built without CONFIG_DEBUG_PAGEALLOC.
826 Note: to get most of debug_pagealloc error reports, it's
827 useful to also enable the page_owner functionality.
828 on: enable the feature
830 debugpat [X86] Enable PAT debugging
832 decnet.addr= [HW,NET]
833 Format: <area>[,<node>]
834 See also Documentation/networking/decnet.txt.
837 [same as hugepagesz=] The size of the default
838 HugeTLB page size. This is the size represented by
839 the legacy /proc/ hugepages APIs, used for SHM, and
840 default size when mounting hugetlbfs filesystems.
841 Defaults to the default architecture's huge page size
844 deferred_probe_timeout=
845 [KNL] Debugging option to set a timeout in seconds for
846 deferred probe to give up waiting on dependencies to
847 probe. Only specific dependencies (subsystems or
848 drivers) that have opted in will be ignored. A timeout of 0
849 will timeout at the end of initcalls. This option will also
850 dump out devices still on the deferred probe list after
854 Format: { on | off | def_only | inf_only | always }
855 on: s390 zlib hardware support for compression on
856 level 1 and decompression (default)
857 off: No s390 zlib hardware support
858 def_only: s390 zlib hardware support for deflate
859 only (compression on level 1)
860 inf_only: s390 zlib hardware support for inflate
862 always: Same as 'on' but ignores the selected compression
863 level always using hardware support (used for debugging)
866 Set number of hash buckets for dentry cache.
868 disable_1tb_segments [PPC]
869 Disables the use of 1TB hash page table segments. This
870 causes the kernel to fall back to 256MB segments which
871 can be useful when debugging issues that require an SLB
875 See Documentation/networking/ipv6.txt.
878 [KNL] Under CONFIG_HARDENED_USERCOPY, whether
879 hardening is enabled for this boot. Hardened
880 usercopy checking is used to protect the kernel
881 from reading or writing beyond known memory
882 allocation boundaries as a proactive defense
883 against bounds-checking flaws in the kernel's
884 copy_to_user()/copy_from_user() interface.
885 on Perform hardened usercopy checks (default).
886 off Disable hardened usercopy checks.
889 Disable RADIX MMU mode on POWER9
892 Disable TLBIE instruction. Currently does not work
893 with KVM, with HASH MMU, or with coherent accelerators.
895 disable_cpu_apicid= [X86,APIC,SMP]
897 The number of initial APIC ID for the
898 corresponding CPU to be disabled at boot,
899 mostly used for the kdump 2nd kernel to
900 disable BSP to wake up multiple CPUs without
901 causing system reset or hang due to sending
904 perf_v4_pmi= [X86,INTEL]
906 Disable Intel PMU counter freezing feature.
907 The feature only exists starting from
908 Arch Perfmon v4 (Skylake and newer).
910 disable_ddw [PPC/PSERIES]
911 Disable Dynamic DMA Window support. Use this if
912 to workaround buggy firmware.
915 See Documentation/networking/ipv6.txt.
917 disable_mtrr_cleanup [X86]
918 The kernel tries to adjust MTRR layout from continuous
919 to discrete, to make X server driver able to add WB
920 entry later. This parameter disables that.
922 disable_mtrr_trim [X86, Intel and AMD only]
923 By default the kernel will trim any uncacheable
924 memory out of your available memory pool based on
925 MTRR settings. This parameter disables that behavior,
926 possibly causing your machine to run very slowly.
928 disable_timer_pin_1 [X86]
929 Disable PIN 1 of APIC timer
930 Can be useful to work around chipset bugs.
932 dis_ucode_ldr [X86] Disable the microcode loader.
934 dma_debug=off If the kernel is compiled with DMA_API_DEBUG support,
935 this option disables the debugging code at boot.
937 dma_debug_entries=<number>
938 This option allows to tune the number of preallocated
939 entries for DMA-API debugging code. One entry is
940 required per DMA-API allocation. Use this if the
941 DMA-API debugging code disables itself because the
942 architectural default is too low.
944 dma_debug_driver=<driver_name>
945 With this option the DMA-API debugging driver
946 filter feature can be enabled at boot time. Just
947 pass the driver to filter for as the parameter.
948 The filter can be disabled or changed to another
949 driver later using sysfs.
951 driver_async_probe= [KNL]
952 List of driver names to be probed asynchronously.
953 Format: <driver_name1>,<driver_name2>...
955 drm.edid_firmware=[<connector>:]<file>[,[<connector>:]<file>]
956 Broken monitors, graphic adapters, KVMs and EDIDless
957 panels may send no or incorrect EDID data sets.
958 This parameter allows to specify an EDID data sets
959 in the /lib/firmware directory that are used instead.
960 Generic built-in EDID data sets are used, if one of
961 edid/1024x768.bin, edid/1280x1024.bin,
962 edid/1680x1050.bin, or edid/1920x1080.bin is given
963 and no file with the same name exists. Details and
964 instructions how to build your own EDID data are
965 available in Documentation/admin-guide/edid.rst. An EDID
966 data set will only be used for a particular connector,
967 if its name and a colon are prepended to the EDID
968 name. Each connector may use a unique EDID data
969 set by separating the files with a comma. An EDID
970 data set with no connector name will be used for
971 any connectors not explicitly specified.
976 Format: {"off" | "known"}
977 Control how the dt_cpu_ftrs device-tree binding is
978 used for CPU feature discovery and setup (if it
980 off: Do not use it, fall back to legacy cpu table.
981 known: Do not pass through unknown features to guests
982 or userspace, only those that the kernel is aware of.
984 dump_apple_properties [X86]
985 Dump name and content of EFI device properties on
986 x86 Macs. Useful for driver authors to determine
987 what data is available or for reverse-engineering.
989 dyndbg[="val"] [KNL,DYNAMIC_DEBUG]
990 module.dyndbg[="val"]
991 Enable debug messages at boot time. See
992 Documentation/admin-guide/dynamic-debug-howto.rst
995 nopku [X86] Disable Memory Protection Keys CPU feature found
998 module.async_probe [KNL]
999 Enable asynchronous probe on this module.
1001 early_ioremap_debug [KNL]
1002 Enable debug messages in early_ioremap support. This
1003 is useful for tracking down temporary early mappings
1004 which are not unmapped.
1006 earlycon= [KNL] Output early console device and options.
1008 When used with no options, the early console is
1009 determined by stdout-path property in device tree's
1010 chosen node or the ACPI SPCR table if supported by
1013 cdns,<addr>[,options]
1014 Start an early, polled-mode console on a Cadence
1015 (xuartps) serial port at the specified address. Only
1016 supported option is baud rate. If baud rate is not
1017 specified, the serial port must already be setup and
1020 uart[8250],io,<addr>[,options]
1021 uart[8250],mmio,<addr>[,options]
1022 uart[8250],mmio32,<addr>[,options]
1023 uart[8250],mmio32be,<addr>[,options]
1024 uart[8250],0x<addr>[,options]
1025 Start an early, polled-mode console on the 8250/16550
1026 UART at the specified I/O port or MMIO address.
1027 MMIO inter-register address stride is either 8-bit
1028 (mmio) or 32-bit (mmio32 or mmio32be).
1029 If none of [io|mmio|mmio32|mmio32be], <addr> is assumed
1030 to be equivalent to 'mmio'. 'options' are specified
1031 in the same format described for "console=ttyS<n>"; if
1032 unspecified, the h/w is not initialized.
1036 Start an early, polled-mode console on a pl011 serial
1037 port at the specified address. The pl011 serial port
1038 must already be setup and configured. Options are not
1039 yet supported. If 'mmio32' is specified, then only
1040 the driver will use only 32-bit accessors to read/write
1041 the device registers.
1044 Start an early, polled-mode console on a meson serial
1045 port at the specified address. The serial port must
1046 already be setup and configured. Options are not yet
1050 Start an early, polled-mode console on an msm serial
1051 port at the specified address. The serial port
1052 must already be setup and configured. Options are not
1055 msm_serial_dm,<addr>
1056 Start an early, polled-mode console on an msm serial
1057 dm port at the specified address. The serial port
1058 must already be setup and configured. Options are not
1062 Start an early, polled-mode console on a serial port
1063 of an Actions Semi SoC, such as S500 or S900, at the
1064 specified address. The serial port must already be
1065 setup and configured. Options are not yet supported.
1068 Start an early, polled-mode console on a serial port
1069 of an RDA Micro SoC, such as RDA8810PL, at the
1070 specified address. The serial port must already be
1071 setup and configured. Options are not yet supported.
1074 Use RISC-V SBI (Supervisor Binary Interface) for early
1077 smh Use ARM semihosting calls for early console.
1085 Use early console provided by serial driver available
1086 on Samsung SoCs, requires selecting proper type and
1087 a correct base address of the selected UART port. The
1088 serial port must already be setup and configured.
1089 Options are not yet supported.
1092 Start an early, polled-mode console on a lantiq serial
1093 (lqasc) port at the specified address. The serial port
1094 must already be setup and configured. Options are not
1099 Use early console provided by Freescale LP UART driver
1100 found on Freescale Vybrid and QorIQ LS1021A processors.
1101 A valid base address must be provided, and the serial
1102 port must already be setup and configured.
1106 Start an early, polled-mode, output-only console on the
1107 Freescale i.MX UART at the specified address. The UART
1108 must already be setup and configured.
1111 Start an early, polled-mode console on the
1112 Armada 3700 serial port at the specified
1113 address. The serial port must already be setup
1114 and configured. Options are not yet supported.
1117 Start an early, polled-mode console on a Qualcomm
1118 Generic Interface (GENI) based serial port at the
1119 specified address. The serial port must already be
1120 setup and configured. Options are not yet supported.
1123 Start an early, unaccelerated console on the EFI
1124 memory mapped framebuffer (if available). On cache
1125 coherent non-x86 systems that use system memory for
1126 the framebuffer, pass the 'ram' option so that it is
1127 mapped with the correct attributes.
1130 Use early console provided by Freescale LINFlexD UART
1131 serial driver for NXP S32V234 SoCs. A valid base
1132 address must be provided, and the serial port must
1133 already be setup and configured.
1135 earlyprintk= [X86,SH,ARM,M68k,S390]
1139 earlyprintk=serial[,ttySn[,baudrate]]
1140 earlyprintk=serial[,0x...[,baudrate]]
1141 earlyprintk=ttySn[,baudrate]
1142 earlyprintk=dbgp[debugController#]
1143 earlyprintk=pciserial[,force],bus:device.function[,baudrate]
1144 earlyprintk=xdbc[xhciController#]
1146 earlyprintk is useful when the kernel crashes before
1147 the normal console is initialized. It is not enabled by
1148 default because it has some cosmetic problems.
1150 Append ",keep" to not disable it when the real console
1153 Only one of vga, efi, serial, or usb debug port can
1156 Currently only ttyS0 and ttyS1 may be specified by
1157 name. Other I/O ports may be explicitly specified
1158 on some architectures (x86 and arm at least) by
1159 replacing ttySn with an I/O port address, like this:
1160 earlyprintk=serial,0x1008,115200
1161 You can find the port for a given device in
1162 /proc/tty/driver/serial:
1163 2: uart:ST16650V2 port:00001008 irq:18 ...
1165 Interaction with the standard serial driver is not
1168 The VGA and EFI output is eventually overwritten by
1171 The xen output can only be used by Xen PV guests.
1173 The sclp output can only be used on s390.
1175 The optional "force" to "pciserial" enables use of a
1176 PCI device even when its classcode is not of the
1179 edac_report= [HW,EDAC] Control how to report EDAC event
1180 Format: {"on" | "off" | "force"}
1181 on: enable EDAC to report H/W event. May be overridden
1182 by other higher priority error reporting module.
1183 off: disable H/W event reporting through EDAC.
1184 force: enforce the use of EDAC to report H/W event.
1187 ekgdboc= [X86,KGDB] Allow early kernel console debugging
1190 This is designed to be used in conjunction with
1191 the boot argument: earlyprintk=vga
1194 Format: {"off" | "on" | "skip[mbr]"}
1197 Format: { "old_map", "nochunk", "noruntime", "debug",
1198 "nosoftreserve", "disable_early_pci_dma",
1199 "no_disable_early_pci_dma" }
1200 old_map [X86-64]: switch to the old ioremap-based EFI
1201 runtime services mapping. [Needs CONFIG_X86_UV=y]
1202 nochunk: disable reading files in "chunks" in the EFI
1203 boot stub, as chunking can cause problems with some
1204 firmware implementations.
1205 noruntime : disable EFI runtime services support
1206 debug: enable misc debug output
1207 nosoftreserve: The EFI_MEMORY_SP (Specific Purpose)
1208 attribute may cause the kernel to reserve the
1209 memory range for a memory mapping driver to
1210 claim. Specify efi=nosoftreserve to disable this
1211 reservation and treat the memory by its base type
1212 (i.e. EFI_CONVENTIONAL_MEMORY / "System RAM").
1213 disable_early_pci_dma: Disable the busmaster bit on all
1214 PCI bridges while in the EFI boot stub
1215 no_disable_early_pci_dma: Leave the busmaster bit set
1216 on all PCI bridges while in the EFI boot stub
1218 efi_no_storage_paranoia [EFI; X86]
1219 Using this parameter you can use more than 50% of
1220 your efi variable storage. Use this parameter only if
1221 you are really sure that your UEFI does sane gc and
1222 fulfills the spec otherwise your board may brick.
1224 efi_fake_mem= nn[KMG]@ss[KMG]:aa[,nn[KMG]@ss[KMG]:aa,..] [EFI; X86]
1225 Add arbitrary attribute to specific memory range by
1226 updating original EFI memory map.
1227 Region of memory which aa attribute is added to is
1230 If efi_fake_mem=2G@4G:0x10000,2G@0x10a0000000:0x10000
1231 is specified, EFI_MEMORY_MORE_RELIABLE(0x10000)
1232 attribute is added to range 0x100000000-0x180000000 and
1233 0x10a0000000-0x1120000000.
1235 If efi_fake_mem=8G@9G:0x40000 is specified, the
1236 EFI_MEMORY_SP(0x40000) attribute is added to
1237 range 0x240000000-0x43fffffff.
1239 Using this parameter you can do debugging of EFI memmap
1240 related features. For example, you can do debugging of
1241 Address Range Mirroring feature even if your box
1242 doesn't support it, or mark specific memory as
1245 efivar_ssdt= [EFI; X86] Name of an EFI variable that contains an SSDT
1246 that is to be dynamically loaded by Linux. If there are
1247 multiple variables with the same name but with different
1248 vendor GUIDs, all of them will be loaded. See
1249 Documentation/admin-guide/acpi/ssdt-overlays.rst for details.
1252 eisa_irq_edge= [PARISC,HW]
1253 See header of drivers/parisc/eisa.c.
1256 See comment before function elanfreq_setup() in
1257 arch/x86/kernel/cpu/cpufreq/elanfreq.c.
1259 elfcorehdr=[size[KMG]@]offset[KMG] [IA64,PPC,SH,X86,S390]
1260 Specifies physical address of start of kernel core
1261 image elf header and optionally the size. Generally
1262 kexec loader will pass this option to capture kernel.
1263 See Documentation/admin-guide/kdump/kdump.rst for details.
1265 enable_mtrr_cleanup [X86]
1266 The kernel tries to adjust MTRR layout from continuous
1267 to discrete, to make X server driver able to add WB
1268 entry later. This parameter enables that.
1270 enable_timer_pin_1 [X86]
1271 Enable PIN 1 of APIC timer
1272 Can be useful to work around chipset bugs
1273 (in particular on some ATI chipsets).
1274 The kernel tries to set a reasonable default.
1276 enforcing [SELINUX] Set initial enforcing status.
1278 See security/selinux/Kconfig help text.
1279 0 -- permissive (log only, no denials).
1280 1 -- enforcing (deny and log).
1282 Value can be changed at runtime via
1283 /sys/fs/selinux/enforce.
1286 Disable Error Record Serialization Table (ERST)
1289 ether= [HW,NET] Ethernet cards parameters
1290 This option is obsoleted by the "netdev=" option, which
1291 has equivalent usage. See its documentation for details.
1295 Permit 'security.evm' to be updated regardless of
1296 current integrity status.
1300 fail_make_request=[KNL]
1301 General fault injection mechanism.
1302 Format: <interval>,<probability>,<space>,<times>
1303 See also Documentation/fault-injection/.
1306 See Documentation/admin-guide/blockdev/floppy.rst.
1308 force_pal_cache_flush
1309 [IA-64] Avoid check_sal_cache_flush which may hang on
1310 buggy SAL_CACHE_FLUSH implementations. Using this
1311 parameter will force ia64_sal_cache_flush to call
1312 ia64_pal_cache_flush instead of SAL_CACHE_FLUSH.
1315 Forcefully enable Physical Address Extension (PAE).
1316 Many Pentium M systems disable PAE but may have a
1317 functionally usable PAE implementation.
1318 Warning: use of this parameter will taint the kernel
1319 and may cause unknown problems.
1322 [FTRACE] will set and start the specified tracer
1323 as early as possible in order to facilitate early
1326 ftrace_dump_on_oops[=orig_cpu]
1327 [FTRACE] will dump the trace buffers on oops.
1328 If no parameter is passed, ftrace will dump
1329 buffers of all CPUs, but if you pass orig_cpu, it will
1330 dump only the buffer of the CPU that triggered the
1333 ftrace_filter=[function-list]
1334 [FTRACE] Limit the functions traced by the function
1335 tracer at boot up. function-list is a comma separated
1336 list of functions. This list can be changed at run
1337 time by the set_ftrace_filter file in the debugfs
1340 ftrace_notrace=[function-list]
1341 [FTRACE] Do not trace the functions specified in
1342 function-list. This list can be changed at run time
1343 by the set_ftrace_notrace file in the debugfs
1346 ftrace_graph_filter=[function-list]
1347 [FTRACE] Limit the top level callers functions traced
1348 by the function graph tracer at boot up.
1349 function-list is a comma separated list of functions
1350 that can be changed at run time by the
1351 set_graph_function file in the debugfs tracing directory.
1353 ftrace_graph_notrace=[function-list]
1354 [FTRACE] Do not trace from the functions specified in
1355 function-list. This list is a comma separated list of
1356 functions that can be changed at run time by the
1357 set_graph_notrace file in the debugfs tracing directory.
1359 ftrace_graph_max_depth=<uint>
1360 [FTRACE] Used with the function graph tracer. This is
1361 the max depth it will trace into a function. This value
1362 can be changed at run time by the max_graph_depth file
1363 in the tracefs tracing directory. default: 0 (no limit)
1365 fw_devlink= [KNL] Create device links between consumer and supplier
1366 devices by scanning the firmware to infer the
1367 consumer/supplier relationships. This feature is
1368 especially useful when drivers are loaded as modules as
1369 it ensures proper ordering of tasks like device probing
1370 (suppliers first, then consumers), supplier boot state
1371 clean up (only after all consumers have probed),
1372 suspend/resume & runtime PM (consumers first, then
1374 Format: { off | permissive | on | rpm }
1375 off -- Don't create device links from firmware info.
1376 permissive -- Create device links from firmware info
1377 but use it only for ordering boot state clean
1378 up (sync_state() calls).
1379 on -- Create device links from firmware info and use it
1380 to enforce probe and suspend/resume ordering.
1381 rpm -- Like "on", but also use to order runtime PM.
1384 [HW,JOY] Multisystem joystick and NES/SNES/PSX pad
1385 support via parallel port (up to 5 devices per port)
1386 Format: <port#>,<pad1>,<pad2>,<pad3>,<pad4>,<pad5>
1387 See also Documentation/input/devices/joystick-parport.rst
1391 gart_fix_e820= [X86_64] disable the fix e820 for K8 GART
1395 gcov_persist= [GCOV] When non-zero (default), profiling data for
1396 kernel modules is saved and remains accessible via
1397 debugfs, even when the module is unloaded/reloaded.
1398 When zero, profiling data is discarded and associated
1399 debugfs files are removed at module unload time.
1401 goldfish [X86] Enable the goldfish android emulator platform.
1402 Don't use this when you are not running on the
1405 gpt [EFI] Forces disk with valid GPT signature but
1406 invalid Protective MBR to be treated as GPT. If the
1407 primary GPT is corrupted, it enables the backup/alternate
1408 GPT to be used instead.
1410 grcan.enable0= [HW] Configuration of physical interface 0. Determines
1411 the "Enable 0" bit of the configuration register.
1414 grcan.enable1= [HW] Configuration of physical interface 1. Determines
1415 the "Enable 0" bit of the configuration register.
1418 grcan.select= [HW] Select which physical interface to use.
1421 grcan.txsize= [HW] Sets the size of the tx buffer.
1422 Format: <unsigned int> such that (txsize & ~0x1fffc0) == 0.
1424 grcan.rxsize= [HW] Sets the size of the rx buffer.
1425 Format: <unsigned int> such that (rxsize & ~0x1fffc0) == 0.
1428 gpio-mockup.gpio_mockup_ranges
1429 [HW] Sets the ranges of gpiochip of for this device.
1430 Format: <start1>,<end1>,<start2>,<end2>...
1432 hardlockup_all_cpu_backtrace=
1433 [KNL] Should the hard-lockup detector generate
1434 backtraces on all cpus.
1437 hashdist= [KNL,NUMA] Large hashes allocated during boot
1438 are distributed across NUMA nodes. Defaults on
1439 for 64-bit NUMA, off otherwise.
1440 Format: 0 | 1 (for off | on)
1442 hcl= [IA-64] SGI's Hardware Graph compatibility layer
1444 hd= [EIDE] (E)IDE hard drive subsystem geometry
1445 Format: <cyl>,<head>,<sect>
1448 Disable Hardware Error Source Table (HEST) support;
1449 corresponding firmware-first mode error processing
1450 logic will be disabled.
1452 highmem=nn[KMG] [KNL,BOOT] forces the highmem zone to have an exact
1453 size of <nn>. This works even on boxes that have no
1454 highmem otherwise. This also works to reduce highmem
1455 size on bigger boxes.
1457 highres= [KNL] Enable/disable high resolution timer mode.
1458 Valid parameters: "on", "off"
1463 hpet= [X86-32,HPET] option to control HPET usage
1464 Format: { enable (default) | disable | force |
1466 disable: disable HPET and use PIT instead
1467 force: allow force enabled of undocumented chips (ICH4,
1469 verbose: show contents of HPET registers during setup
1471 hpet_mmap= [X86, HPET_MMAP] Allow userspace to mmap HPET
1472 registers. Default set by CONFIG_HPET_MMAP_DEFAULT.
1474 hugetlb_cma= [HW] The size of a cma area used for allocation
1475 of gigantic hugepages.
1478 Reserve a cma area of given size and allocate gigantic
1479 hugepages using the cma allocator. If enabled, the
1480 boot-time allocation of gigantic hugepages is skipped.
1482 hugepages= [HW,X86-32,IA-64] HugeTLB pages to allocate at boot.
1483 hugepagesz= [HW,IA-64,PPC,X86-64] The size of the HugeTLB pages.
1484 On x86-64 and powerpc, this option can be specified
1485 multiple times interleaved with hugepages= to reserve
1486 huge pages of different sizes. Valid pages sizes on
1487 x86-64 are 2M (when the CPU supports "pse") and 1G
1488 (when the CPU supports the "pdpe1gb" cpuinfo flag).
1491 [KNL] Should the hung task detector generate panics.
1494 A nonzero value instructs the kernel to panic when a
1495 hung task is detected. The default value is controlled
1496 by the CONFIG_BOOTPARAM_HUNG_TASK_PANIC build-time
1497 option. The value selected by this boot parameter can
1498 be changed later by the kernel.hung_task_panic sysctl.
1500 hvc_iucv= [S390] Number of z/VM IUCV hypervisor console (HVC)
1501 terminal devices. Valid values: 0..8
1502 hvc_iucv_allow= [S390] Comma-separated list of z/VM user IDs.
1503 If specified, z/VM IUCV HVC accepts connections
1504 from listed z/VM user IDs only.
1506 hv_nopvspin [X86,HYPER_V] Disables the paravirt spinlock optimizations
1507 which allow the hypervisor to 'idle' the
1508 guest on lock contention.
1511 Do not unregister boot console at start. This is only
1512 useful for debugging when something happens in the window
1513 between unregistering the boot console and initializing
1516 i2c_bus= [HW] Override the default board specific I2C bus speed
1517 or register an additional I2C bus that is not
1518 registered from board initialization code.
1522 i8042.debug [HW] Toggle i8042 debug mode
1523 i8042.unmask_kbd_data
1524 [HW] Enable printing of interrupt data from the KBD port
1525 (disabled by default, and as a pre-condition
1526 requires that i8042.debug=1 be enabled)
1527 i8042.direct [HW] Put keyboard port into non-translated mode
1528 i8042.dumbkbd [HW] Pretend that controller can only read data from
1529 keyboard and cannot control its state
1530 (Don't attempt to blink the leds)
1531 i8042.noaux [HW] Don't check for auxiliary (== mouse) port
1532 i8042.nokbd [HW] Don't check/create keyboard port
1533 i8042.noloop [HW] Disable the AUX Loopback command while probing
1535 i8042.nomux [HW] Don't check presence of an active multiplexing
1537 i8042.nopnp [HW] Don't use ACPIPnP / PnPBIOS to discover KBD/AUX
1539 i8042.notimeout [HW] Ignore timeout condition signalled by controller
1540 i8042.reset [HW] Reset the controller during init, cleanup and
1541 suspend-to-ram transitions, only during s2r
1542 transitions, or never reset
1543 Format: { 1 | Y | y | 0 | N | n }
1544 1, Y, y: always reset controller
1545 0, N, n: don't ever reset controller
1546 Default: only on s2r transitions on x86; most other
1547 architectures force reset to be always executed
1548 i8042.unlock [HW] Unlock (ignore) the keylock
1549 i8042.kbdreset [HW] Reset device connected to KBD port
1553 i8k.ignore_dmi [HW] Continue probing hardware even if DMI data
1554 indicates that the driver is running on unsupported
1556 i8k.force [HW] Activate i8k driver even if SMM BIOS signature
1557 does not match list of supported models.
1559 [HW] Report power status in /proc/i8k
1560 (disabled by default)
1561 i8k.restricted [HW] Allow controlling fans only if SYS_ADMIN
1564 i915.invert_brightness=
1565 [DRM] Invert the sense of the variable that is used to
1566 set the brightness of the panel backlight. Normally a
1567 brightness value of 0 indicates backlight switched off,
1568 and the maximum of the brightness value sets the backlight
1569 to maximum brightness. If this parameter is set to 0
1570 (default) and the machine requires it, or this parameter
1571 is set to 1, a brightness value of 0 sets the backlight
1572 to maximum brightness, and the maximum of the brightness
1573 value switches the backlight off.
1574 -1 -- never invert brightness
1575 0 -- machine default
1576 1 -- force brightness inversion
1579 Format: <io>[,<membase>[,<icn_id>[,<icn_id2>]]]
1581 ide-core.nodma= [HW] (E)IDE subsystem
1582 Format: =0.0 to prevent dma on hda, =0.1 hdb =1.0 hdc
1583 .vlb_clock .pci_clock .noflush .nohpa .noprobe .nowerr
1584 .cdrom .chs .ignore_cable are additional options
1585 See Documentation/ide/ide.rst.
1587 ide-generic.probe-mask= [HW] (E)IDE subsystem
1589 Probe mask for legacy ISA IDE ports. Depending on
1590 platform up to 6 ports are supported, enabled by
1591 setting corresponding bits in the mask to 1. The
1592 default value is 0x0, which has a special meaning.
1593 On systems that have PCI, it triggers scanning the
1594 PCI bus for the first and the second port, which
1595 are then probed. On systems without PCI the value
1596 of 0x0 enables probing the two first ports as if it
1599 ide-pci-generic.all-generic-ide [HW] (E)IDE subsystem
1600 Claim all unknown PCI IDE storage controllers.
1603 Format: idle=poll, idle=halt, idle=nomwait
1604 Poll forces a polling idle loop that can slightly
1605 improve the performance of waking up a idle CPU, but
1606 will use a lot of power and make the system run hot.
1608 idle=halt: Halt is forced to be used for CPU idle.
1609 In such case C2/C3 won't be used again.
1610 idle=nomwait: Disable mwait for CPU C-states
1612 ieee754= [MIPS] Select IEEE Std 754 conformance mode
1613 Format: { strict | legacy | 2008 | relaxed }
1616 Choose which programs will be accepted for execution
1617 based on the IEEE 754 NaN encoding(s) supported by
1618 the FPU and the NaN encoding requested with the value
1619 of an ELF file header flag individually set by each
1620 binary. Hardware implementations are permitted to
1621 support either or both of the legacy and the 2008 NaN
1624 Available settings are as follows:
1625 strict accept binaries that request a NaN encoding
1626 supported by the FPU
1627 legacy only accept legacy-NaN binaries, if supported
1629 2008 only accept 2008-NaN binaries, if supported
1631 relaxed accept any binaries regardless of whether
1632 supported by the FPU
1634 The FPU emulator is always able to support both NaN
1635 encodings, so if no FPU hardware is present or it has
1636 been disabled with 'nofpu', then the settings of
1637 'legacy' and '2008' strap the emulator accordingly,
1638 'relaxed' straps the emulator for both legacy-NaN and
1639 2008-NaN, whereas 'strict' enables legacy-NaN only on
1640 legacy processors and both NaN encodings on MIPS32 or
1643 The setting for ABS.fmt/NEG.fmt instruction execution
1644 mode generally follows that for the NaN encoding,
1645 except where unsupported by hardware.
1647 ignore_loglevel [KNL]
1648 Ignore loglevel setting - this will print /all/
1649 kernel messages to the console. Useful for debugging.
1650 We also add it as printk module parameter, so users
1651 could change it dynamically, usually by
1652 /sys/module/printk/parameters/ignore_loglevel.
1655 Ignore RLIMIT_DATA setting for data mappings,
1656 print warning at first misuse. Can be changed via
1657 /sys/module/kernel/parameters/ignore_rlimit_data.
1659 ihash_entries= [KNL]
1660 Set number of hash buckets for inode cache.
1662 ima_appraise= [IMA] appraise integrity measurements
1663 Format: { "off" | "enforce" | "fix" | "log" }
1666 ima_appraise_tcb [IMA] Deprecated. Use ima_policy= instead.
1667 The builtin appraise policy appraises all files
1670 ima_canonical_fmt [IMA]
1671 Use the canonical format for the binary runtime
1672 measurements, instead of host native format.
1675 Format: { md5 | sha1 | rmd160 | sha256 | sha384
1679 The list of supported hash algorithms is defined
1680 in crypto/hash_info.h.
1683 The builtin policies to load during IMA setup.
1684 Format: "tcb | appraise_tcb | secure_boot |
1687 The "tcb" policy measures all programs exec'd, files
1688 mmap'd for exec, and all files opened with the read
1689 mode bit set by either the effective uid (euid=0) or
1692 The "appraise_tcb" policy appraises the integrity of
1693 all files owned by root.
1695 The "secure_boot" policy appraises the integrity
1696 of files (eg. kexec kernel image, kernel modules,
1697 firmware, policy, etc) based on file signatures.
1699 The "fail_securely" policy forces file signature
1700 verification failure also on privileged mounted
1701 filesystems with the SB_I_UNVERIFIABLE_SIGNATURE
1704 ima_tcb [IMA] Deprecated. Use ima_policy= instead.
1705 Load a policy which meets the needs of the Trusted
1706 Computing Base. This means IMA will measure all
1707 programs exec'd, files mmap'd for exec, and all files
1708 opened for read by uid=0.
1711 Select one of defined IMA measurements template formats.
1712 Formats: { "ima" | "ima-ng" | "ima-sig" }
1716 [IMA] Define a custom template format.
1717 Format: { "field1|...|fieldN" }
1719 ima.ahash_minsize= [IMA] Minimum file size for asynchronous hash usage
1720 Format: <min_file_size>
1721 Set the minimal file size for using asynchronous hash.
1722 If left unspecified, ahash usage is disabled.
1724 ahash performance varies for different data sizes on
1725 different crypto accelerators. This option can be used
1726 to achieve the best performance for a particular HW.
1728 ima.ahash_bufsize= [IMA] Asynchronous hash buffer size
1730 Set hashing buffer size. Default: 4k.
1732 ahash performance varies for different chunk sizes on
1733 different crypto accelerators. This option can be used
1734 to achieve best performance for particular HW.
1738 Run specified binary instead of /sbin/init as init
1741 initcall_debug [KNL] Trace initcalls as they are executed. Useful
1742 for working out where the kernel is dying during
1745 initcall_blacklist= [KNL] Do not execute a comma-separated list of
1746 initcall functions. Useful for debugging built-in
1747 modules and initcalls.
1749 initrd= [BOOT] Specify the location of the initial ramdisk
1751 initrdmem= [KNL] Specify a physical address and size from which to
1752 load the initrd. If an initrd is compiled in or
1753 specified in the bootparams, it takes priority over this
1755 Format: ss[KMG],nn[KMG]
1758 init_on_alloc= [MM] Fill newly allocated pages and heap objects with
1761 Default set by CONFIG_INIT_ON_ALLOC_DEFAULT_ON.
1763 init_on_free= [MM] Fill freed pages and heap objects with zeroes.
1765 Default set by CONFIG_INIT_ON_FREE_DEFAULT_ON.
1767 init_pkru= [x86] Specify the default memory protection keys rights
1768 register contents for all processes. 0x55555554 by
1769 default (disallow access to all but pkey 0). Can
1770 override in debugfs after boot.
1772 inport.irq= [HW] Inport (ATI XL and Microsoft) busmouse driver
1775 int_pln_enable [x86] Enable power limit notification interrupt
1777 integrity_audit=[IMA]
1778 Format: { "0" | "1" }
1779 0 -- basic integrity auditing messages. (Default)
1780 1 -- additional integrity auditing messages.
1782 intel_iommu= [DMAR] Intel IOMMU driver (DMAR) option
1784 Enable intel iommu driver.
1786 Disable intel iommu driver.
1787 igfx_off [Default Off]
1788 By default, gfx is mapped as normal device. If a gfx
1789 device has a dedicated DMAR unit, the DMAR unit is
1790 bypassed by not enabling DMAR with this option. In
1791 this case, gfx device will use physical address for
1794 With this option iommu will not optimize to look
1795 for io virtual address below 32-bit forcing dual
1796 address cycle on pci bus for cards supporting greater
1797 than 32-bit addressing. The default is to look
1798 for translation below 32-bit and if not available
1799 then look in the higher range.
1800 strict [Default Off]
1801 With this option on every unmap_single operation will
1802 result in a hardware IOTLB flush operation as opposed
1803 to batching them for performance.
1804 sp_off [Default Off]
1805 By default, super page will be supported if Intel IOMMU
1806 has the capability. With this option, super page will
1809 By default, scalable mode will be disabled even if the
1810 hardware advertises that it has support for the scalable
1811 mode translation. With this option set, scalable mode
1812 will be used on hardware which claims to support it.
1813 tboot_noforce [Default Off]
1814 Do not force the Intel IOMMU enabled under tboot.
1815 By default, tboot will force Intel IOMMU on, which
1816 could harm performance of some high-throughput
1817 devices like 40GBit network cards, even if identity
1819 Note that using this option lowers the security
1820 provided by tboot because it makes the system
1821 vulnerable to DMA attacks.
1822 nobounce [Default off]
1823 Disable bounce buffer for untrusted devices such as
1824 the Thunderbolt devices. This will treat the untrusted
1825 devices as the trusted ones, hence might expose security
1826 risks of DMA attacks.
1828 intel_idle.max_cstate= [KNL,HW,ACPI,X86]
1829 0 disables intel_idle and fall back on acpi_idle.
1830 1 to 9 specify maximum depth of C-state.
1834 Do not enable intel_pstate as the default
1835 scaling driver for the supported processors
1837 Use intel_pstate as a scaling driver, but configure it
1838 to work with generic cpufreq governors (instead of
1839 enabling its internal governor). This mode cannot be
1840 used along with the hardware-managed P-states (HWP)
1843 Enable intel_pstate on systems that prohibit it by default
1844 in favor of acpi-cpufreq. Forcing the intel_pstate driver
1845 instead of acpi-cpufreq may disable platform features, such
1846 as thermal controls and power capping, that rely on ACPI
1847 P-States information being indicated to OSPM and therefore
1848 should be used with caution. This option does not work with
1849 processors that aren't supported by the intel_pstate driver
1850 or on platforms that use pcc-cpufreq instead of acpi-cpufreq.
1852 Do not enable hardware P state control (HWP)
1855 Only load intel_pstate on systems which support
1856 hardware P state control (HWP) if available.
1858 Enforce ACPI _PPC performance limits. If the Fixed ACPI
1859 Description Table, specifies preferred power management
1860 profile as "Enterprise Server" or "Performance Server",
1861 then this feature is turned on by default.
1863 Allow per-logical-CPU P-State performance control limits using
1864 cpufreq sysfs interface
1866 intremap= [X86-64, Intel-IOMMU]
1867 on enable Interrupt Remapping (default)
1868 off disable Interrupt Remapping
1869 nosid disable Source ID checking
1871 BIOS x2APIC opt-out request will be ignored
1872 nopost disable Interrupt Posting
1874 iomem= Disable strict checking of access to MMIO memory
1875 strict regions from userspace.
1890 nobypass [PPC/POWERNV]
1891 Disable IOMMU bypass, using IOMMU for PCI devices.
1893 iommu.strict= [ARM64] Configure TLB invalidation behaviour
1894 Format: { "0" | "1" }
1896 Request that DMA unmap operations use deferred
1897 invalidation of hardware TLBs, for increased
1898 throughput at the cost of reduced device isolation.
1899 Will fall back to strict mode if not supported by
1900 the relevant IOMMU driver.
1901 1 - Strict mode (default).
1902 DMA unmap operations invalidate IOMMU hardware TLBs
1906 [ARM64, X86] Configure DMA to bypass the IOMMU by default.
1907 Format: { "0" | "1" }
1908 0 - Use IOMMU translation for DMA.
1909 1 - Bypass the IOMMU for DMA.
1910 unset - Use value of CONFIG_IOMMU_DEFAULT_PASSTHROUGH.
1912 io7= [HW] IO7 for Marvel based alpha systems
1913 See comment before marvel_specify_io7 in
1914 arch/alpha/kernel/core_marvel.c.
1916 io_delay= [X86] I/O delay method
1918 Standard port 0x80 based delay
1920 Alternate port 0xed based delay (needed on some systems)
1922 Simple two microseconds delay
1927 See Documentation/admin-guide/nfs/nfsroot.rst.
1929 ipcmni_extend [KNL] Extend the maximum number of unique System V
1930 IPC identifiers from 32,768 to 16,777,216.
1932 irqaffinity= [SMP] Set the default irq affinity mask
1933 The argument is a cpu list, as described above.
1935 irqchip.gicv2_force_probe=
1938 Force the kernel to look for the second 4kB page
1939 of a GICv2 controller even if the memory range
1940 exposed by the device tree is too small.
1942 irqchip.gicv3_nolpi=
1944 Force the kernel to ignore the availability of
1945 LPIs (and by consequence ITSs). Intended for system
1946 that use the kernel as a bootloader, and thus want
1947 to let secondary kernels in charge of setting up
1950 irqchip.gicv3_pseudo_nmi= [ARM64]
1951 Enables support for pseudo-NMIs in the kernel. This
1952 requires the kernel to be built with
1953 CONFIG_ARM64_PSEUDO_NMI.
1956 When an interrupt is not handled search all handlers
1957 for it. Intended to get systems with badly broken
1961 When an interrupt is not handled search all handlers
1962 for it. Also check all handlers each timer
1963 interrupt. Intended to get systems with badly broken
1967 Format: <RDP>,<reset>,<pci_scan>,<verbosity>
1969 isolcpus= [KNL,SMP,ISOL] Isolate a given set of CPUs from disturbance.
1970 [Deprecated - use cpusets instead]
1971 Format: [flag-list,]<cpu-list>
1973 Specify one or more CPUs to isolate from disturbances
1974 specified in the flag list (default: domain):
1977 Disable the tick when a single task runs.
1979 A residual 1Hz tick is offloaded to workqueues, which you
1980 need to affine to housekeeping through the global
1981 workqueue's affinity configured via the
1982 /sys/devices/virtual/workqueue/cpumask sysfs file, or
1983 by using the 'domain' flag described below.
1985 NOTE: by default the global workqueue runs on all CPUs,
1986 so to protect individual CPUs the 'cpumask' file has to
1987 be configured manually after bootup.
1990 Isolate from the general SMP balancing and scheduling
1991 algorithms. Note that performing domain isolation this way
1992 is irreversible: it's not possible to bring back a CPU to
1993 the domains once isolated through isolcpus. It's strongly
1994 advised to use cpusets instead to disable scheduler load
1995 balancing through the "cpuset.sched_load_balance" file.
1996 It offers a much more flexible interface where CPUs can
1997 move in and out of an isolated set anytime.
1999 You can move a process onto or off an "isolated" CPU via
2000 the CPU affinity syscalls or cpuset.
2001 <cpu number> begins at 0 and the maximum value is
2002 "number of CPUs in system - 1".
2006 Isolate from being targeted by managed interrupts
2007 which have an interrupt mask containing isolated
2008 CPUs. The affinity of managed interrupts is
2009 handled by the kernel and cannot be changed via
2010 the /proc/irq/* interfaces.
2012 This isolation is best effort and only effective
2013 if the automatically assigned interrupt mask of a
2014 device queue contains isolated and housekeeping
2015 CPUs. If housekeeping CPUs are online then such
2016 interrupts are directed to the housekeeping CPU
2017 so that IO submitted on the housekeeping CPU
2018 cannot disturb the isolated CPU.
2020 If a queue's affinity mask contains only isolated
2021 CPUs then this parameter has no effect on the
2022 interrupt routing decision, though interrupts are
2023 only delivered when tasks running on those
2024 isolated CPUs submit IO. IO submitted on
2025 housekeeping CPUs has no influence on those
2028 The format of <cpu-list> is described above.
2032 ivrs_ioapic [HW,X86_64]
2033 Provide an override to the IOAPIC-ID<->DEVICE-ID
2034 mapping provided in the IVRS ACPI table. For
2035 example, to map IOAPIC-ID decimal 10 to
2036 PCI device 00:14.0 write the parameter as:
2037 ivrs_ioapic[10]=00:14.0
2039 ivrs_hpet [HW,X86_64]
2040 Provide an override to the HPET-ID<->DEVICE-ID
2041 mapping provided in the IVRS ACPI table. For
2042 example, to map HPET-ID decimal 0 to
2043 PCI device 00:14.0 write the parameter as:
2044 ivrs_hpet[0]=00:14.0
2046 ivrs_acpihid [HW,X86_64]
2047 Provide an override to the ACPI-HID:UID<->DEVICE-ID
2048 mapping provided in the IVRS ACPI table. For
2049 example, to map UART-HID:UID AMD0020:0 to
2050 PCI device 00:14.5 write the parameter as:
2051 ivrs_acpihid[00:14.5]=AMD0020:0
2053 js= [HW,JOY] Analog joystick
2054 See Documentation/input/joydev/joystick.rst.
2057 When CONFIG_RANDOMIZE_BASE is set, this disables
2058 kernel and module base offset ASLR (Address Space
2059 Layout Randomization).
2062 [KNL] Enforce KASAN (Kernel Address Sanitizer) to print
2063 report on every invalid memory access. Without this
2064 parameter KASAN will print report only for the first
2069 kernelcore= [KNL,X86,IA-64,PPC]
2070 Format: nn[KMGTPE] | nn% | "mirror"
2071 This parameter specifies the amount of memory usable by
2072 the kernel for non-movable allocations. The requested
2073 amount is spread evenly throughout all nodes in the
2074 system as ZONE_NORMAL. The remaining memory is used for
2075 movable memory in its own zone, ZONE_MOVABLE. In the
2076 event, a node is too small to have both ZONE_NORMAL and
2077 ZONE_MOVABLE, kernelcore memory will take priority and
2078 other nodes will have a larger ZONE_MOVABLE.
2080 ZONE_MOVABLE is used for the allocation of pages that
2081 may be reclaimed or moved by the page migration
2082 subsystem. Note that allocations like PTEs-from-HighMem
2083 still use the HighMem zone if it exists, and the Normal
2084 zone if it does not.
2086 It is possible to specify the exact amount of memory in
2087 the form of "nn[KMGTPE]", a percentage of total system
2088 memory in the form of "nn%", or "mirror". If "mirror"
2089 option is specified, mirrored (reliable) memory is used
2090 for non-movable allocations and remaining memory is used
2091 for Movable pages. "nn[KMGTPE]", "nn%", and "mirror"
2092 are exclusive, so you cannot specify multiple forms.
2094 kgdbdbgp= [KGDB,HW] kgdb over EHCI usb debug port.
2095 Format: <Controller#>[,poll interval]
2096 The controller # is the number of the ehci usb debug
2097 port as it is probed via PCI. The poll interval is
2098 optional and is the number seconds in between
2099 each poll cycle to the debug port in case you need
2100 the functionality for interrupting the kernel with
2101 gdb or control-c on the dbgp connection. When
2102 not using this parameter you use sysrq-g to break into
2103 the kernel debugger.
2105 kgdboc= [KGDB,HW] kgdb over consoles.
2106 Requires a tty driver that supports console polling,
2107 or a supported polling keyboard driver (non-usb).
2108 Serial only format: <serial_device>[,baud]
2109 keyboard only format: kbd
2110 keyboard and serial format: kbd,<serial_device>[,baud]
2111 Optional Kernel mode setting:
2112 kms, kbd format: kms,kbd
2113 kms, kbd and serial format: kms,kbd,<ser_dev>[,baud]
2115 kgdbwait [KGDB] Stop kernel execution and enter the
2116 kernel debugger at the earliest opportunity.
2118 kmac= [MIPS] korina ethernet MAC address.
2119 Configure the RouterBoard 532 series on-chip
2120 Ethernet adapter MAC address.
2122 kmemleak= [KNL] Boot-time kmemleak enable/disable
2123 Valid arguments: on, off
2125 Built with CONFIG_DEBUG_KMEMLEAK_DEFAULT_OFF=y,
2128 kprobe_event=[probe-list]
2129 [FTRACE] Add kprobe events and enable at boot time.
2130 The probe-list is a semicolon delimited list of probe
2131 definitions. Each definition is same as kprobe_events
2132 interface, but the parameters are comma delimited.
2133 For example, to add a kprobe event on vfs_read with
2134 arg1 and arg2, add to the command line;
2136 kprobe_event=p,vfs_read,$arg1,$arg2
2138 See also Documentation/trace/kprobetrace.rst "Kernel
2139 Boot Parameter" section.
2141 kpti= [ARM64] Control page table isolation of user
2142 and kernel address spaces.
2143 Default: enabled on cores which need mitigation.
2147 kvm.ignore_msrs=[KVM] Ignore guest accesses to unhandled MSRs.
2148 Default is 0 (don't ignore, but inject #GP)
2150 kvm.enable_vmware_backdoor=[KVM] Support VMware backdoor PV interface.
2151 Default is false (don't support).
2153 kvm.mmu_audit= [KVM] This is a R/W parameter which allows audit
2158 [KVM] Controls the software workaround for the
2159 X86_BUG_ITLB_MULTIHIT bug.
2160 force : Always deploy workaround.
2161 off : Never deploy workaround.
2162 auto : Deploy workaround based on the presence of
2163 X86_BUG_ITLB_MULTIHIT.
2167 If the software workaround is enabled for the host,
2168 guests do need not to enable it for nested guests.
2170 kvm.nx_huge_pages_recovery_ratio=
2171 [KVM] Controls how many 4KiB pages are periodically zapped
2172 back to huge pages. 0 disables the recovery, otherwise if
2173 the value is N KVM will zap 1/Nth of the 4KiB pages every
2174 minute. The default is 60.
2176 kvm-amd.nested= [KVM,AMD] Allow nested virtualization in KVM/SVM.
2177 Default is 1 (enabled)
2179 kvm-amd.npt= [KVM,AMD] Disable nested paging (virtualized MMU)
2181 Default is 1 (enabled) if in 64-bit or 32-bit PAE mode.
2183 kvm-arm.vgic_v3_group0_trap=
2184 [KVM,ARM] Trap guest accesses to GICv3 group-0
2187 kvm-arm.vgic_v3_group1_trap=
2188 [KVM,ARM] Trap guest accesses to GICv3 group-1
2191 kvm-arm.vgic_v3_common_trap=
2192 [KVM,ARM] Trap guest accesses to GICv3 common
2195 kvm-arm.vgic_v4_enable=
2196 [KVM,ARM] Allow use of GICv4 for direct injection of
2199 kvm-intel.ept= [KVM,Intel] Disable extended page tables
2200 (virtualized MMU) support on capable Intel chips.
2201 Default is 1 (enabled)
2203 kvm-intel.emulate_invalid_guest_state=
2204 [KVM,Intel] Enable emulation of invalid guest states
2205 Default is 0 (disabled)
2207 kvm-intel.flexpriority=
2208 [KVM,Intel] Disable FlexPriority feature (TPR shadow).
2209 Default is 1 (enabled)
2212 [KVM,Intel] Enable VMX nesting (nVMX).
2213 Default is 0 (disabled)
2215 kvm-intel.unrestricted_guest=
2216 [KVM,Intel] Disable unrestricted guest feature
2217 (virtualized real and unpaged mode) on capable
2218 Intel chips. Default is 1 (enabled)
2220 kvm-intel.vmentry_l1d_flush=[KVM,Intel] Mitigation for L1 Terminal Fault
2223 Valid arguments: never, cond, always
2225 always: L1D cache flush on every VMENTER.
2226 cond: Flush L1D on VMENTER only when the code between
2227 VMEXIT and VMENTER can leak host memory.
2228 never: Disables the mitigation
2230 Default is cond (do L1 cache flush in specific instances)
2232 kvm-intel.vpid= [KVM,Intel] Disable Virtual Processor Identification
2233 feature (tagged TLBs) on capable Intel chips.
2234 Default is 1 (enabled)
2236 l1tf= [X86] Control mitigation of the L1TF vulnerability on
2239 The kernel PTE inversion protection is unconditionally
2240 enabled and cannot be disabled.
2243 Provides all available mitigations for the
2244 L1TF vulnerability. Disables SMT and
2245 enables all mitigations in the
2246 hypervisors, i.e. unconditional L1D flush.
2248 SMT control and L1D flush control via the
2249 sysfs interface is still possible after
2250 boot. Hypervisors will issue a warning
2251 when the first VM is started in a
2252 potentially insecure configuration,
2253 i.e. SMT enabled or L1D flush disabled.
2256 Same as 'full', but disables SMT and L1D
2257 flush runtime control. Implies the
2258 'nosmt=force' command line option.
2259 (i.e. sysfs control of SMT is disabled.)
2262 Leaves SMT enabled and enables the default
2263 hypervisor mitigation, i.e. conditional
2266 SMT control and L1D flush control via the
2267 sysfs interface is still possible after
2268 boot. Hypervisors will issue a warning
2269 when the first VM is started in a
2270 potentially insecure configuration,
2271 i.e. SMT enabled or L1D flush disabled.
2275 Disables SMT and enables the default
2276 hypervisor mitigation.
2278 SMT control and L1D flush control via the
2279 sysfs interface is still possible after
2280 boot. Hypervisors will issue a warning
2281 when the first VM is started in a
2282 potentially insecure configuration,
2283 i.e. SMT enabled or L1D flush disabled.
2286 Same as 'flush', but hypervisors will not
2287 warn when a VM is started in a potentially
2288 insecure configuration.
2291 Disables hypervisor mitigations and doesn't
2293 It also drops the swap size and available
2294 RAM limit restriction on both hypervisor and
2299 For details see: Documentation/admin-guide/hw-vuln/l1tf.rst
2305 lapic [X86-32,APIC] Enable the local APIC even if BIOS
2308 lapic= [x86,APIC] "notscdeadline" Do not use TSC deadline
2309 value for LAPIC timer one-shot implementation. Default
2310 back to the programmable timer unit in the LAPIC.
2312 lapic_timer_c2_ok [X86,APIC] trust the local apic timer
2315 libata.dma= [LIBATA] DMA control
2316 libata.dma=0 Disable all PATA and SATA DMA
2317 libata.dma=1 PATA and SATA Disk DMA only
2318 libata.dma=2 ATAPI (CDROM) DMA only
2319 libata.dma=4 Compact Flash DMA only
2320 Combinations also work, so libata.dma=3 enables DMA
2321 for disks and CDROMs, but not CFs.
2323 libata.ignore_hpa= [LIBATA] Ignore HPA limit
2324 libata.ignore_hpa=0 keep BIOS limits (default)
2325 libata.ignore_hpa=1 ignore limits, using full disk
2327 libata.noacpi [LIBATA] Disables use of ACPI in libata suspend/resume
2331 libata.force= [LIBATA] Force configurations. The format is comma
2332 separated list of "[ID:]VAL" where ID is
2333 PORT[.DEVICE]. PORT and DEVICE are decimal numbers
2334 matching port, link or device. Basically, it matches
2335 the ATA ID string printed on console by libata. If
2336 the whole ID part is omitted, the last PORT and DEVICE
2337 values are used. If ID hasn't been specified yet, the
2338 configuration applies to all ports, links and devices.
2340 If only DEVICE is omitted, the parameter applies to
2341 the port and all links and devices behind it. DEVICE
2342 number of 0 either selects the first device or the
2343 first fan-out link behind PMP device. It does not
2344 select the host link. DEVICE number of 15 selects the
2345 host link and device attached to it.
2347 The VAL specifies the configuration to force. As long
2348 as there's no ambiguity shortcut notation is allowed.
2349 For example, both 1.5 and 1.5G would work for 1.5Gbps.
2350 The following configurations can be forced.
2352 * Cable type: 40c, 80c, short40c, unk, ign or sata.
2353 Any ID with matching PORT is used.
2355 * SATA link speed limit: 1.5Gbps or 3.0Gbps.
2357 * Transfer mode: pio[0-7], mwdma[0-4] and udma[0-7].
2358 udma[/][16,25,33,44,66,100,133] notation is also
2361 * [no]ncq: Turn on or off NCQ.
2363 * [no]ncqtrim: Turn off queued DSM TRIM.
2365 * nohrst, nosrst, norst: suppress hard, soft
2368 * rstonce: only attempt one reset during
2369 hot-unplug link recovery
2371 * dump_id: dump IDENTIFY data.
2373 * atapi_dmadir: Enable ATAPI DMADIR bridge support
2375 * disable: Disable this device.
2377 If there are multiple matching configurations changing
2378 the same attribute, the last one is used.
2380 memblock=debug [KNL] Enable memblock debug messages.
2382 load_ramdisk= [RAM] List of ramdisks to load from floppy
2383 See Documentation/admin-guide/blockdev/ramdisk.rst.
2385 lockd.nlm_grace_period=P [NFS] Assign grace period.
2388 lockd.nlm_tcpport=N [NFS] Assign TCP port.
2391 lockd.nlm_timeout=T [NFS] Assign timeout value.
2394 lockd.nlm_udpport=M [NFS] Assign UDP port.
2397 lockdown= [SECURITY]
2398 { integrity | confidentiality }
2399 Enable the kernel lockdown feature. If set to
2400 integrity, kernel features that allow userland to
2401 modify the running kernel are disabled. If set to
2402 confidentiality, kernel features that allow userland
2403 to extract confidential information from the kernel
2406 locktorture.nreaders_stress= [KNL]
2407 Set the number of locking read-acquisition kthreads.
2408 Defaults to being automatically set based on the
2409 number of online CPUs.
2411 locktorture.nwriters_stress= [KNL]
2412 Set the number of locking write-acquisition kthreads.
2414 locktorture.onoff_holdoff= [KNL]
2415 Set time (s) after boot for CPU-hotplug testing.
2417 locktorture.onoff_interval= [KNL]
2418 Set time (s) between CPU-hotplug operations, or
2419 zero to disable CPU-hotplug testing.
2421 locktorture.shuffle_interval= [KNL]
2422 Set task-shuffle interval (jiffies). Shuffling
2423 tasks allows some CPUs to go into dyntick-idle
2424 mode during the locktorture test.
2426 locktorture.shutdown_secs= [KNL]
2427 Set time (s) after boot system shutdown. This
2428 is useful for hands-off automated testing.
2430 locktorture.stat_interval= [KNL]
2431 Time (s) between statistics printk()s.
2433 locktorture.stutter= [KNL]
2434 Time (s) to stutter testing, for example,
2435 specifying five seconds causes the test to run for
2436 five seconds, wait for five seconds, and so on.
2437 This tests the locking primitive's ability to
2438 transition abruptly to and from idle.
2440 locktorture.torture_type= [KNL]
2441 Specify the locking implementation to test.
2443 locktorture.verbose= [KNL]
2444 Enable additional printk() statements.
2446 logibm.irq= [HW,MOUSE] Logitech Bus Mouse Driver
2449 loglevel= All Kernel Messages with a loglevel smaller than the
2450 console loglevel will be printed to the console. It can
2451 also be changed with klogd or other programs. The
2452 loglevels are defined as follows:
2454 0 (KERN_EMERG) system is unusable
2455 1 (KERN_ALERT) action must be taken immediately
2456 2 (KERN_CRIT) critical conditions
2457 3 (KERN_ERR) error conditions
2458 4 (KERN_WARNING) warning conditions
2459 5 (KERN_NOTICE) normal but significant condition
2460 6 (KERN_INFO) informational
2461 7 (KERN_DEBUG) debug-level messages
2463 log_buf_len=n[KMG] Sets the size of the printk ring buffer,
2464 in bytes. n must be a power of two and greater
2465 than the minimal size. The minimal size is defined
2466 by LOG_BUF_SHIFT kernel config parameter. There is
2467 also CONFIG_LOG_CPU_MAX_BUF_SHIFT config parameter
2468 that allows to increase the default size depending on
2469 the number of CPUs. See init/Kconfig for more details.
2471 logo.nologo [FB] Disables display of the built-in Linux logo.
2472 This may be used to provide more screen space for
2473 kernel log messages and is useful when debugging
2474 kernel boot problems.
2476 lp=0 [LP] Specify parallel ports to use, e.g,
2477 lp=port[,port...] lp=none,parport0 (lp0 not configured, lp1 uses
2478 lp=reset first parallel port). 'lp=0' disables the
2479 lp=auto printer driver. 'lp=reset' (which can be
2480 specified in addition to the ports) causes
2481 attached printers to be reset. Using
2482 lp=port1,port2,... specifies the parallel ports
2483 to associate lp devices with, starting with
2484 lp0. A port specification may be 'none' to skip
2485 that lp device, or a parport name such as
2486 'parport0'. Specifying 'lp=auto' instead of a
2487 port specification list means that device IDs
2488 from each port should be examined, to see if
2489 an IEEE 1284-compliant printer is attached; if
2490 so, the driver will manage that printer.
2491 See also header of drivers/char/lp.c.
2494 Sets loops_per_jiffy to given constant, thus avoiding
2495 time-consuming boot-time autodetection (up to 250 ms per
2496 CPU). 0 enables autodetection (default). To determine
2497 the correct value for your kernel, boot with normal
2498 autodetection and see what value is printed. Note that
2499 on SMP systems the preset will be applied to all CPUs,
2500 which is likely to cause problems if your CPUs need
2501 significantly divergent settings. An incorrect value
2502 will cause delays in the kernel to be wrong, leading to
2503 unpredictable I/O errors and other breakage. Although
2504 unlikely, in the extreme case this might damage your
2508 Format: <io>,<irq>,<dma>
2510 lsm.debug [SECURITY] Enable LSM initialization debugging output.
2513 [SECURITY] Choose order of LSM initialization. This
2514 overrides CONFIG_LSM, and the "security=" parameter.
2516 machvec= [IA-64] Force the use of a particular machine-vector
2517 (machvec) in a generic kernel.
2518 Example: machvec=hpzx1
2520 machtype= [Loongson] Share the same kernel image file between different
2522 Example: machtype=lemote-yeeloong-2f-7inch
2524 max_addr=nn[KMG] [KNL,BOOT,ia64] All physical memory greater
2525 than or equal to this physical address is ignored.
2527 maxcpus= [SMP] Maximum number of processors that an SMP kernel
2528 will bring up during bootup. maxcpus=n : n >= 0 limits
2529 the kernel to bring up 'n' processors. Surely after
2530 bootup you can bring up the other plugged cpu by executing
2531 "echo 1 > /sys/devices/system/cpu/cpuX/online". So maxcpus
2532 only takes effect during system bootup.
2533 While n=0 is a special case, it is equivalent to "nosmp",
2534 which also disables the IO APIC.
2536 max_loop= [LOOP] The number of loop block devices that get
2537 (loop.max_loop) unconditionally pre-created at init time. The default
2538 number is configured by BLK_DEV_LOOP_MIN_COUNT. Instead
2539 of statically allocating a predefined number, loop
2540 devices can be requested on-demand with the
2541 /dev/loop-control interface.
2543 mce [X86-32] Machine Check Exception
2545 mce=option [X86-64] See Documentation/x86/x86_64/boot-options.rst
2547 md= [HW] RAID subsystems devices and level
2548 See Documentation/admin-guide/md.rst.
2551 Format: <first>,<last>
2552 Specifies range of consoles to be captured by the MDA.
2555 Control mitigation for the Micro-architectural Data
2556 Sampling (MDS) vulnerability.
2558 Certain CPUs are vulnerable to an exploit against CPU
2559 internal buffers which can forward information to a
2560 disclosure gadget under certain conditions.
2562 In vulnerable processors, the speculatively
2563 forwarded data can be used in a cache side channel
2564 attack, to access data to which the attacker does
2565 not have direct access.
2567 This parameter controls the MDS mitigation. The
2570 full - Enable MDS mitigation on vulnerable CPUs
2571 full,nosmt - Enable MDS mitigation and disable
2572 SMT on vulnerable CPUs
2573 off - Unconditionally disable MDS mitigation
2575 On TAA-affected machines, mds=off can be prevented by
2576 an active TAA mitigation as both vulnerabilities are
2577 mitigated with the same mechanism so in order to disable
2578 this mitigation, you need to specify tsx_async_abort=off
2581 Not specifying this option is equivalent to
2584 For details see: Documentation/admin-guide/hw-vuln/mds.rst
2586 mem=nn[KMG] [KNL,BOOT] Force usage of a specific amount of memory
2587 Amount of memory to be used in cases as follows:
2590 2 when the kernel is not able to see the whole system memory;
2591 3 memory that lies after 'mem=' boundary is excluded from
2592 the hypervisor, then assigned to KVM guests.
2594 [X86] Work as limiting max address. Use together
2595 with memmap= to avoid physical address space collisions.
2596 Without memmap= PCI devices could be placed at addresses
2597 belonging to unused RAM.
2599 Note that this only takes effects during boot time since
2600 in above case 3, memory may need be hot added after boot
2601 if system memory of hypervisor is not sufficient.
2603 mem=nopentium [BUGS=X86-32] Disable usage of 4MB pages for kernel
2607 [KNL,SH] Allow user to override the default size for
2608 per-device physically contiguous DMA buffers.
2610 memhp_default_state=online/offline
2611 [KNL] Set the initial state for the memory hotplug
2612 onlining policy. If not specified, the default value is
2613 set according to the
2614 CONFIG_MEMORY_HOTPLUG_DEFAULT_ONLINE kernel config
2616 See Documentation/admin-guide/mm/memory-hotplug.rst.
2618 memmap=exactmap [KNL,X86] Enable setting of an exact
2619 E820 memory map, as specified by the user.
2620 Such memmap=exactmap lines can be constructed based on
2621 BIOS output or other requirements. See the memmap=nn@ss
2624 memmap=nn[KMG]@ss[KMG]
2625 [KNL] Force usage of a specific region of memory.
2626 Region of memory to be used is from ss to ss+nn.
2627 If @ss[KMG] is omitted, it is equivalent to mem=nn[KMG],
2628 which limits max address to nn[KMG].
2629 Multiple different regions can be specified,
2632 memmap=100M@2G,100M#3G,1G!1024G
2634 memmap=nn[KMG]#ss[KMG]
2635 [KNL,ACPI] Mark specific memory as ACPI data.
2636 Region of memory to be marked is from ss to ss+nn.
2638 memmap=nn[KMG]$ss[KMG]
2639 [KNL,ACPI] Mark specific memory as reserved.
2640 Region of memory to be reserved is from ss to ss+nn.
2641 Example: Exclude memory from 0x18690000-0x1869ffff
2642 memmap=64K$0x18690000
2644 memmap=0x10000$0x18690000
2645 Some bootloaders may need an escape character before '$',
2646 like Grub2, otherwise '$' and the following number
2649 memmap=nn[KMG]!ss[KMG]
2650 [KNL,X86] Mark specific memory as protected.
2651 Region of memory to be used, from ss to ss+nn.
2652 The memory region may be marked as e820 type 12 (0xc)
2653 and is NVDIMM or ADR memory.
2655 memmap=<size>%<offset>-<oldtype>+<newtype>
2656 [KNL,ACPI] Convert memory within the specified region
2657 from <oldtype> to <newtype>. If "-<oldtype>" is left
2658 out, the whole region will be marked as <newtype>,
2659 even if previously unavailable. If "+<newtype>" is left
2660 out, matching memory will be removed. Types are
2661 specified as e820 types, e.g., 1 = RAM, 2 = reserved,
2662 3 = ACPI, 12 = PRAM.
2664 memory_corruption_check=0/1 [X86]
2665 Some BIOSes seem to corrupt the first 64k of
2666 memory when doing things like suspend/resume.
2667 Setting this option will scan the memory
2668 looking for corruption. Enabling this will
2669 both detect corruption and prevent the kernel
2670 from using the memory being corrupted.
2671 However, its intended as a diagnostic tool; if
2672 repeatable BIOS-originated corruption always
2673 affects the same memory, you can use memmap=
2674 to prevent the kernel from using that memory.
2676 memory_corruption_check_size=size [X86]
2677 By default it checks for corruption in the low
2678 64k, making this memory unavailable for normal
2679 use. Use this parameter to scan for
2680 corruption in more or less memory.
2682 memory_corruption_check_period=seconds [X86]
2683 By default it checks for corruption every 60
2684 seconds. Use this parameter to check at some
2685 other rate. 0 disables periodic checking.
2687 memtest= [KNL,X86,ARM,PPC] Enable memtest
2689 default : 0 <disable>
2690 Specifies the number of memtest passes to be
2691 performed. Each pass selects another test
2692 pattern from a given set of patterns. Memtest
2693 fills the memory with this pattern, validates
2694 memory contents and reserves bad memory
2695 regions that are detected.
2697 mem_encrypt= [X86-64] AMD Secure Memory Encryption (SME) control
2698 Valid arguments: on, off
2699 Default (depends on kernel configuration option):
2700 on (CONFIG_AMD_MEM_ENCRYPT_ACTIVE_BY_DEFAULT=y)
2701 off (CONFIG_AMD_MEM_ENCRYPT_ACTIVE_BY_DEFAULT=n)
2702 mem_encrypt=on: Activate SME
2703 mem_encrypt=off: Do not activate SME
2705 Refer to Documentation/virt/kvm/amd-memory-encryption.rst
2706 for details on when memory encryption can be activated.
2708 mem_sleep_default= [SUSPEND] Default system suspend mode:
2709 s2idle - Suspend-To-Idle
2710 shallow - Power-On Suspend or equivalent (if supported)
2711 deep - Suspend-To-RAM or equivalent (if supported)
2712 See Documentation/admin-guide/pm/sleep-states.rst.
2714 meye.*= [HW] Set MotionEye Camera parameters
2715 See Documentation/media/v4l-drivers/meye.rst.
2717 mfgpt_irq= [IA-32] Specify the IRQ to use for the
2718 Multi-Function General Purpose Timers on AMD Geode
2721 mfgptfix [X86-32] Fix MFGPT timers on AMD Geode platforms when
2722 the BIOS has incorrectly applied a workaround. TinyBIOS
2723 version 0.98 is known to be affected, 0.99 fixes the
2724 problem by letting the user disable the workaround.
2728 min_addr=nn[KMG] [KNL,BOOT,ia64] All physical memory below this
2729 physical address is ignored.
2731 mini2440= [ARM,HW,KNL]
2732 Format:[0..2][b][c][t]
2734 MINI2440 configuration specification:
2735 0 - The attached screen is the 3.5" TFT
2736 1 - The attached screen is the 7" TFT
2737 2 - The VGA Shield is attached (1024x768)
2738 Leaving out the screen size parameter will not load
2739 the TFT driver, and the framebuffer will be left
2741 b - Enable backlight. The TFT backlight pin will be
2742 linked to the kernel VESA blanking code and a GPIO
2743 LED. This parameter is not necessary when using the
2745 c - Enable the s3c camera interface.
2746 t - Reserved for enabling touchscreen support. The
2747 touchscreen support is not enabled in the mainstream
2748 kernel as of 2.6.30, a preliminary port can be found
2749 in the "bleeding edge" mini2440 support kernel at
2750 http://repo.or.cz/w/linux-2.6/mini2440.git
2753 [X86,PPC,S390,ARM64] Control optional mitigations for
2754 CPU vulnerabilities. This is a set of curated,
2755 arch-independent options, each of which is an
2756 aggregation of existing arch-specific options.
2759 Disable all optional CPU mitigations. This
2760 improves system performance, but it may also
2761 expose users to several CPU vulnerabilities.
2762 Equivalent to: nopti [X86,PPC]
2764 nospectre_v1 [X86,PPC]
2766 nospectre_v2 [X86,PPC,S390,ARM64]
2767 spectre_v2_user=off [X86]
2768 spec_store_bypass_disable=off [X86,PPC]
2769 ssbd=force-off [ARM64]
2772 tsx_async_abort=off [X86]
2773 kvm.nx_huge_pages=off [X86]
2776 This does not have any effect on
2777 kvm.nx_huge_pages when
2778 kvm.nx_huge_pages=force.
2781 Mitigate all CPU vulnerabilities, but leave SMT
2782 enabled, even if it's vulnerable. This is for
2783 users who don't want to be surprised by SMT
2784 getting disabled across kernel upgrades, or who
2785 have other ways of avoiding SMT-based attacks.
2786 Equivalent to: (default behavior)
2789 Mitigate all CPU vulnerabilities, disabling SMT
2790 if needed. This is for users who always want to
2791 be fully mitigated, even if it means losing SMT.
2792 Equivalent to: l1tf=flush,nosmt [X86]
2793 mds=full,nosmt [X86]
2794 tsx_async_abort=full,nosmt [X86]
2797 [KNL] When CONFIG_DEBUG_MEMORY_INIT is set, this
2798 parameter allows control of the logging verbosity for
2799 the additional memory initialisation checks. A value
2800 of 0 disables mminit logging and a level of 4 will
2801 log everything. Information is printed at KERN_DEBUG
2802 so loglevel=8 may also need to be specified.
2805 [KNL] When CONFIG_MODULE_SIG is set, this means that
2806 modules without (valid) signatures will fail to load.
2807 Note that if CONFIG_MODULE_SIG_FORCE is set, that
2808 is always true, so this option does nothing.
2810 module_blacklist= [KNL] Do not load a comma-separated list of
2811 modules. Useful for debugging problem modules.
2814 [MOUSE] Maximum time between finger touching and
2815 leaving touchpad surface for touch to be considered
2816 a tap and be reported as a left button click (for
2817 touchpads working in absolute mode only).
2819 mousedev.xres= [MOUSE] Horizontal screen resolution, used for devices
2820 reporting absolute coordinates, such as tablets
2821 mousedev.yres= [MOUSE] Vertical screen resolution, used for devices
2822 reporting absolute coordinates, such as tablets
2824 movablecore= [KNL,X86,IA-64,PPC]
2825 Format: nn[KMGTPE] | nn%
2826 This parameter is the complement to kernelcore=, it
2827 specifies the amount of memory used for migratable
2828 allocations. If both kernelcore and movablecore is
2829 specified, then kernelcore will be at *least* the
2830 specified value but may be more. If movablecore on its
2831 own is specified, the administrator must be careful
2832 that the amount of memory usable for all allocations
2835 movable_node [KNL] Boot-time switch to make hotplugable memory
2836 NUMA nodes to be movable. This means that the memory
2837 of such nodes will be usable only for movable
2838 allocations which rules out almost all kernel
2839 allocations. Use with caution!
2841 MTD_Partition= [MTD]
2842 Format: <name>,<region-number>,<size>,<offset>
2844 MTD_Region= [MTD] Format:
2845 <name>,<region-number>[,<base>,<size>,<buswidth>,<altbuswidth>]
2848 See drivers/mtd/parsers/cmdlinepart.c
2850 multitce=off [PPC] This parameter disables the use of the pSeries
2851 firmware feature for updating multiple TCE entries
2854 onenand.bdry= [HW,MTD] Flex-OneNAND Boundary Configuration
2856 Format: [die0_boundary][,die0_lock][,die1_boundary][,die1_lock]
2858 boundary - index of last SLC block on Flex-OneNAND.
2859 The remaining blocks are configured as MLC blocks.
2860 lock - Configure if Flex-OneNAND boundary should be locked.
2861 Once locked, the boundary cannot be changed.
2862 1 indicates lock status, 0 indicates unlock status.
2865 ARM/S3C2412 JIVE boot control
2867 See arch/arm/mach-s3c2412/mach-jive.c
2869 mtouchusb.raw_coordinates=
2870 [HW] Make the MicroTouch USB driver use raw coordinates
2871 ('y', default) or cooked coordinates ('n')
2873 mtrr_chunk_size=nn[KMG] [X86]
2874 used for mtrr cleanup. It is largest continuous chunk
2875 that could hold holes aka. UC entries.
2877 mtrr_gran_size=nn[KMG] [X86]
2878 Used for mtrr cleanup. It is granularity of mtrr block.
2880 Large value could prevent small alignment from
2883 mtrr_spare_reg_nr=n [X86]
2885 Range: 0,7 : spare reg number
2887 Used for mtrr cleanup. It is spare mtrr entries number.
2888 Set to 2 or more if your graphical card needs more.
2890 n2= [NET] SDL Inc. RISCom/N2 synchronous serial card
2892 netdev= [NET] Network devices parameters
2893 Format: <irq>,<io>,<mem_start>,<mem_end>,<name>
2894 Note that mem_start is often overloaded to mean
2895 something different and driver-specific.
2896 This usage is only documented in each driver source
2900 [NETFILTER] Enable connection tracking flow accounting
2901 0 to disable accounting
2902 1 to enable accounting
2905 nfsaddrs= [NFS] Deprecated. Use ip= instead.
2906 See Documentation/admin-guide/nfs/nfsroot.rst.
2908 nfsroot= [NFS] nfs root filesystem for disk-less boxes.
2909 See Documentation/admin-guide/nfs/nfsroot.rst.
2911 nfsrootdebug [NFS] enable nfsroot debugging messages.
2912 See Documentation/admin-guide/nfs/nfsroot.rst.
2914 nfs.callback_nr_threads=
2915 [NFSv4] set the total number of threads that the
2916 NFS client will assign to service NFSv4 callback
2919 nfs.callback_tcpport=
2920 [NFS] set the TCP port on which the NFSv4 callback
2921 channel should listen.
2924 [NFS] sets the pathname to the program which is used
2925 to update the NFS client cache entries.
2927 nfs.cache_getent_timeout=
2928 [NFS] sets the timeout after which an attempt to
2929 update a cache entry is deemed to have failed.
2931 nfs.idmap_cache_timeout=
2932 [NFS] set the maximum lifetime for idmapper cache
2936 [NFS] enable 64-bit inode numbers.
2937 If zero, the NFS client will fake up a 32-bit inode
2938 number for the readdir() and stat() syscalls instead
2939 of returning the full 64-bit number.
2940 The default is to return 64-bit inode numbers.
2942 nfs.max_session_cb_slots=
2943 [NFSv4.1] Sets the maximum number of session
2944 slots the client will assign to the callback
2945 channel. This determines the maximum number of
2946 callbacks the client will process in parallel for
2947 a particular server.
2949 nfs.max_session_slots=
2950 [NFSv4.1] Sets the maximum number of session slots
2951 the client will attempt to negotiate with the server.
2952 This limits the number of simultaneous RPC requests
2953 that the client can send to the NFSv4.1 server.
2954 Note that there is little point in setting this
2955 value higher than the max_tcp_slot_table_limit.
2957 nfs.nfs4_disable_idmapping=
2958 [NFSv4] When set to the default of '1', this option
2959 ensures that both the RPC level authentication
2960 scheme and the NFS level operations agree to use
2961 numeric uids/gids if the mount is using the
2962 'sec=sys' security flavour. In effect it is
2963 disabling idmapping, which can make migration from
2964 legacy NFSv2/v3 systems to NFSv4 easier.
2965 Servers that do not support this mode of operation
2966 will be autodetected by the client, and it will fall
2967 back to using the idmapper.
2968 To turn off this behaviour, set the value to '0'.
2970 [NFS4] Specify an additional fixed unique ident-
2971 ification string that NFSv4 clients can insert into
2972 their nfs_client_id4 string. This is typically a
2973 UUID that is generated at system install time.
2975 nfs.send_implementation_id =
2976 [NFSv4.1] Send client implementation identification
2977 information in exchange_id requests.
2978 If zero, no implementation identification information
2980 The default is to send the implementation identification
2983 nfs.recover_lost_locks =
2984 [NFSv4] Attempt to recover locks that were lost due
2985 to a lease timeout on the server. Please note that
2986 doing this risks data corruption, since there are
2987 no guarantees that the file will remain unchanged
2988 after the locks are lost.
2989 If you want to enable the kernel legacy behaviour of
2990 attempting to recover these locks, then set this
2992 The default parameter value of '0' causes the kernel
2993 not to attempt recovery of lost locks.
2995 nfs4.layoutstats_timer =
2996 [NFSv4.2] Change the rate at which the kernel sends
2997 layoutstats to the pNFS metadata server.
2999 Setting this to value to 0 causes the kernel to use
3000 whatever value is the default set by the layout
3001 driver. A non-zero value sets the minimum interval
3002 in seconds between layoutstats transmissions.
3004 nfsd.nfs4_disable_idmapping=
3005 [NFSv4] When set to the default of '1', the NFSv4
3006 server will return only numeric uids and gids to
3007 clients using auth_sys, and will accept numeric uids
3008 and gids from such clients. This is intended to ease
3009 migration from NFSv2/v3.
3011 nmi_debug= [KNL,SH] Specify one or more actions to take
3012 when a NMI is triggered.
3013 Format: [state][,regs][,debounce][,die]
3015 nmi_watchdog= [KNL,BUGS=X86] Debugging features for SMP kernels
3016 Format: [panic,][nopanic,][num]
3018 0 - turn hardlockup detector in nmi_watchdog off
3019 1 - turn hardlockup detector in nmi_watchdog on
3020 When panic is specified, panic when an NMI watchdog
3021 timeout occurs (or 'nopanic' to not panic on an NMI
3022 watchdog, if CONFIG_BOOTPARAM_HARDLOCKUP_PANIC is set)
3023 To disable both hard and soft lockup detectors,
3024 please see 'nowatchdog'.
3025 This is useful when you use a panic=... timeout and
3026 need the box quickly up again.
3028 These settings can be accessed at runtime via
3029 the nmi_watchdog and hardlockup_panic sysctls.
3031 netpoll.carrier_timeout=
3032 [NET] Specifies amount of time (in seconds) that
3033 netpoll should wait for a carrier. By default netpoll
3036 no387 [BUGS=X86-32] Tells the kernel to use the 387 maths
3037 emulation library even if a 387 maths coprocessor
3040 no5lvl [X86-64] Disable 5-level paging mode. Forces
3041 kernel to use 4-level paging instead.
3044 [HW] Never suspend the console
3045 Disable suspending of consoles during suspend and
3046 hibernate operations. Once disabled, debugging
3047 messages can reach various consoles while the rest
3048 of the system is being put to sleep (ie, while
3049 debugging driver suspend/resume hooks). This may
3050 not work reliably with all consoles, but is known
3051 to work with serial and VGA consoles.
3052 To facilitate more flexible debugging, we also add
3053 console_suspend, a printk module parameter to control
3054 it. Users could use console_suspend (usually
3055 /sys/module/printk/parameters/console_suspend) to
3056 turn on/off it dynamically.
3058 novmcoredd [KNL,KDUMP]
3059 Disable device dump. Device dump allows drivers to
3060 append dump data to vmcore so you can collect driver
3061 specified debug info. Drivers can append the data
3062 without any limit and this data is stored in memory,
3063 so this may cause significant memory stress. Disabling
3064 device dump can help save memory but the driver debug
3065 data will be no longer available. This parameter
3066 is only available when CONFIG_PROC_VMCORE_DEVICE_DUMP
3069 noaliencache [MM, NUMA, SLAB] Disables the allocation of alien
3070 caches in the slab allocator. Saves per-node memory,
3071 but will impact performance.
3075 noaltinstr [S390] Disables alternative instructions patching
3076 (CPU alternatives feature).
3078 noapic [SMP,APIC] Tells the kernel to not make use of any
3079 IOAPICs that may be present in the system.
3081 noautogroup Disable scheduler automatic task group creation.
3083 nobats [PPC] Do not use BATs for mapping kernel lowmem
3084 on "Classic" PPC cores.
3088 noclflush [BUGS=X86] Don't use the CLFLUSH instruction
3090 nodelayacct [KNL] Disable per-task delay accounting
3092 nodsp [SH] Disable hardware DSP at boot time.
3094 noefi Disable EFI runtime services support.
3099 On X86-32 available only on PAE configured kernels.
3100 noexec=on: enable non-executable mappings (default)
3101 noexec=off: disable non-executable mappings
3104 Disable SMAP (Supervisor Mode Access Prevention)
3105 even if it is supported by processor.
3108 Disable SMEP (Supervisor Mode Execution Prevention)
3109 even if it is supported by processor.
3112 This affects only 32-bit executables.
3113 noexec32=on: enable non-executable mappings (default)
3114 read doesn't imply executable mappings
3115 noexec32=off: disable non-executable mappings
3116 read implies executable mappings
3118 nofpu [MIPS,SH] Disable hardware FPU at boot time.
3120 nofxsr [BUGS=X86-32] Disables x86 floating point extended
3121 register save and restore. The kernel will only save
3122 legacy floating-point registers on task switch.
3124 nohugeiomap [KNL,x86,PPC] Disable kernel huge I/O mappings.
3126 nosmt [KNL,S390] Disable symmetric multithreading (SMT).
3127 Equivalent to smt=1.
3129 [KNL,x86] Disable symmetric multithreading (SMT).
3130 nosmt=force: Force disable SMT, cannot be undone
3131 via the sysfs control file.
3133 nospectre_v1 [X86,PPC] Disable mitigations for Spectre Variant 1
3134 (bounds check bypass). With this option data leaks are
3135 possible in the system.
3137 nospectre_v2 [X86,PPC_FSL_BOOK3E,ARM64] Disable all mitigations for
3138 the Spectre variant 2 (indirect branch prediction)
3139 vulnerability. System may allow data leaks with this
3142 nospec_store_bypass_disable
3143 [HW] Disable all mitigations for the Speculative Store Bypass vulnerability
3145 noxsave [BUGS=X86] Disables x86 extended register state save
3146 and restore using xsave. The kernel will fallback to
3147 enabling legacy floating-point and sse state.
3149 noxsaveopt [X86] Disables xsaveopt used in saving x86 extended
3150 register states. The kernel will fall back to use
3151 xsave to save the states. By using this parameter,
3152 performance of saving the states is degraded because
3153 xsave doesn't support modified optimization while
3154 xsaveopt supports it on xsaveopt enabled systems.
3156 noxsaves [X86] Disables xsaves and xrstors used in saving and
3157 restoring x86 extended register state in compacted
3158 form of xsave area. The kernel will fall back to use
3159 xsaveopt and xrstor to save and restore the states
3160 in standard form of xsave area. By using this
3161 parameter, xsave area per process might occupy more
3162 memory on xsaves enabled systems.
3164 nohlt [BUGS=ARM,SH] Tells the kernel that the sleep(SH) or
3165 wfi(ARM) instruction doesn't work correctly and not to
3166 use it. This is also useful when using JTAG debugger.
3168 no_file_caps Tells the kernel not to honor file capabilities. The
3169 only way then for a file to be executed with privilege
3170 is to be setuid root or executed by root.
3172 nohalt [IA-64] Tells the kernel not to use the power saving
3173 function PAL_HALT_LIGHT when idle. This increases
3174 power-consumption. On the positive side, it reduces
3175 interrupt wake-up latency, which may improve performance
3176 in certain environments such as networked servers or
3179 nohibernate [HIBERNATION] Disable hibernation and resume.
3181 nohz= [KNL] Boottime enable/disable dynamic ticks
3182 Valid arguments: on, off
3185 nohz_full= [KNL,BOOT,SMP,ISOL]
3186 The argument is a cpu list, as described above.
3187 In kernels built with CONFIG_NO_HZ_FULL=y, set
3188 the specified list of CPUs whose tick will be stopped
3189 whenever possible. The boot CPU will be forced outside
3190 the range to maintain the timekeeping. Any CPUs
3191 in this list will have their RCU callbacks offloaded,
3192 just as if they had also been called out in the
3193 rcu_nocbs= boot parameter.
3195 noiotrap [SH] Disables trapped I/O port accesses.
3197 noirqdebug [X86-32] Disables the code which attempts to detect and
3198 disable unhandled interrupt sources.
3200 no_timer_check [X86,APIC] Disables the code which tests for
3201 broken timer IRQ sources.
3203 noisapnp [ISAPNP] Disables ISA PnP code.
3205 noinitrd [RAM] Tells the kernel not to load any configured
3208 nointremap [X86-64, Intel-IOMMU] Do not enable interrupt
3210 [Deprecated - use intremap=off]
3214 noinvpcid [X86] Disable the INVPCID cpu feature.
3216 nojitter [IA-64] Disables jitter checking for ITC timers.
3218 no-kvmclock [X86,KVM] Disable paravirtualized KVM clock driver
3220 no-kvmapf [X86,KVM] Disable paravirtualized asynchronous page
3224 [X86,PV_OPS] Disable paravirtualized VMware scheduler
3225 clock and use the default one.
3227 no-steal-acc [X86,PV_OPS,ARM64] Disable paravirtualized steal time
3228 accounting. steal time is computed, but won't
3229 influence scheduler behaviour
3231 nolapic [X86-32,APIC] Do not enable or use the local APIC.
3233 nolapic_timer [X86-32,APIC] Do not use the local APIC timer.
3235 noltlbs [PPC] Do not use large page/tlb entries for kernel
3236 lowmem mapping on PPC40x and PPC8xx
3238 nomca [IA-64] Disable machine check abort handling
3240 nomce [X86-32] Disable Machine Check Exception
3242 nomfgpt [X86-32] Disable Multi-Function General Purpose
3243 Timer usage (for AMD Geode machines).
3245 nonmi_ipi [X86] Disable using NMI IPIs during panic/reboot to
3246 shutdown the other cpus. Instead use the REBOOT_VECTOR
3249 nomodule Disable module load
3251 nopat [X86] Disable PAT (page attribute table extension of
3252 pagetables) support.
3254 nopcid [X86-64] Disable the PCID cpu feature.
3256 norandmaps Don't use address space randomization. Equivalent to
3257 echo 0 > /proc/sys/kernel/randomize_va_space
3259 noreplace-smp [X86-32,SMP] Don't replace SMP instructions
3260 with UP alternatives
3262 nordrand [X86] Disable kernel use of the RDRAND and
3263 RDSEED instructions even if they are supported
3264 by the processor. RDRAND and RDSEED are still
3265 available to user space applications.
3267 noresume [SWSUSP] Disables resume and restores original swap
3270 no-scroll [VGA] Disables scrollback.
3271 This is required for the Braillex ib80-piezo Braille
3272 reader made by F.H. Papenmeier (Germany).
3276 nosep [BUGS=X86-32] Disables x86 SYSENTER/SYSEXIT support.
3278 nosmp [SMP] Tells an SMP kernel to act as a UP kernel,
3279 and disable the IO APIC. legacy for "maxcpus=0".
3281 nosoftlockup [KNL] Disable the soft-lockup detector.
3283 nosync [HW,M68K] Disables sync negotiation for all devices.
3285 nowatchdog [KNL] Disable both lockup detectors, i.e.
3286 soft-lockup and NMI watchdog (hard-lockup).
3290 nox2apic [X86-64,APIC] Do not enable x2APIC mode.
3292 cpu0_hotplug [X86] Turn on CPU0 hotplug feature when
3293 CONFIG_BOOTPARAM_HOTPLUG_CPU0 is off.
3294 Some features depend on CPU0. Known dependencies are:
3295 1. Resume from suspend/hibernate depends on CPU0.
3296 Suspend/hibernate will fail if CPU0 is offline and you
3297 need to online CPU0 before suspend/hibernate.
3298 2. PIC interrupts also depend on CPU0. CPU0 can't be
3299 removed if a PIC interrupt is detected.
3300 It's said poweroff/reboot may depend on CPU0 on some
3301 machines although I haven't seen such issues so far
3302 after CPU0 is offline on a few tested machines.
3303 If the dependencies are under your control, you can
3304 turn on cpu0_hotplug.
3306 nps_mtm_hs_ctr= [KNL,ARC]
3307 This parameter sets the maximum duration, in
3308 cycles, each HW thread of the CTOP can run
3309 without interruptions, before HW switches it.
3310 The actual maximum duration is 16 times this
3312 Format: integer between 1 and 255
3315 nptcg= [IA-64] Override max number of concurrent global TLB
3316 purges which is reported from either PAL_VM_SUMMARY or
3319 nr_cpus= [SMP] Maximum number of processors that an SMP kernel
3320 could support. nr_cpus=n : n >= 1 limits the kernel to
3321 support 'n' processors. It could be larger than the
3322 number of already plugged CPU during bootup, later in
3323 runtime you can physically add extra cpu until it reaches
3324 n. So during boot up some boot time memory for per-cpu
3325 variables need be pre-allocated for later physical cpu
3328 nr_uarts= [SERIAL] maximum number of UARTs to be registered.
3330 numa_balancing= [KNL,X86] Enable or disable automatic NUMA balancing.
3331 Allowed values are enable and disable
3333 numa_zonelist_order= [KNL, BOOT] Select zonelist order for NUMA.
3334 'node', 'default' can be specified
3335 This can be set from sysctl after boot.
3336 See Documentation/admin-guide/sysctl/vm.rst for details.
3338 ohci1394_dma=early [HW] enable debugging via the ohci1394 driver.
3339 See Documentation/debugging-via-ohci1394.txt for more
3342 olpc_ec_timeout= [OLPC] ms delay when issuing EC commands
3343 Rather than timing out after 20 ms if an EC
3344 command is not properly ACKed, override the length
3345 of the timeout. We have interrupts disabled while
3346 waiting for the ACK, so if this is set too high
3347 interrupts *may* be lost!
3349 omap_mux= [OMAP] Override bootloader pin multiplexing.
3350 Format: <mux_mode0.mode_name=value>...
3351 For example, to override I2C bus2:
3352 omap_mux=i2c2_scl.i2c2_scl=0x100,i2c2_sda.i2c2_sda=0x100
3354 oprofile.timer= [HW]
3355 Use timer interrupt instead of performance counters
3357 oprofile.cpu_type= Force an oprofile cpu type
3358 This might be useful if you have an older oprofile
3359 userland or if you want common events.
3360 Format: { arch_perfmon }
3361 arch_perfmon: [X86] Force use of architectural
3362 perfmon on Intel CPUs instead of the
3363 CPU specific event set.
3364 timer: [X86] Force use of architectural NMI
3365 timer mode (see also oprofile.timer
3366 for generic hr timer mode)
3368 oops=panic Always panic on oopses. Default is to just kill the
3369 process, but there is a small probability of
3370 deadlocking the machine.
3371 This will also cause panics on machine check exceptions.
3372 Useful together with panic=30 to trigger a reboot.
3375 [KNL] Boolean flag to control whether the page allocator
3376 should randomize its free lists. The randomization may
3377 be automatically enabled if the kernel detects it is
3378 running on a platform with a direct-mapped memory-side
3379 cache, and this parameter can be used to
3380 override/disable that behavior. The state of the flag
3381 can be read from sysfs at:
3382 /sys/module/page_alloc/parameters/shuffle.
3384 page_owner= [KNL] Boot-time page_owner enabling option.
3385 Storage of the information about who allocated
3386 each page is disabled in default. With this switch,
3388 on: enable the feature
3390 page_poison= [KNL] Boot-time parameter changing the state of
3391 poisoning on the buddy allocator, available with
3392 CONFIG_PAGE_POISONING=y.
3393 off: turn off poisoning (default)
3394 on: turn on poisoning
3396 panic= [KNL] Kernel behaviour on panic: delay <timeout>
3397 timeout > 0: seconds before rebooting
3398 timeout = 0: wait forever
3399 timeout < 0: reboot immediately
3402 panic_print= Bitmask for printing system info when panic happens.
3403 User can chose combination of the following bits:
3404 bit 0: print all tasks info
3405 bit 1: print system memory info
3406 bit 2: print timer info
3407 bit 3: print locks info if CONFIG_LOCKDEP is on
3408 bit 4: print ftrace buffer
3409 bit 5: print all printk messages in buffer
3411 panic_on_warn panic() instead of WARN(). Useful to cause kdump
3414 crash_kexec_post_notifiers
3415 Run kdump after running panic-notifiers and dumping
3416 kmsg. This only for the users who doubt kdump always
3417 succeeds in any situation.
3418 Note that this also increases risks of kdump failure,
3419 because some panic notifiers can make the crashed
3420 kernel more unstable.
3422 parkbd.port= [HW] Parallel port number the keyboard adapter is
3423 connected to, default is 0.
3425 parkbd.mode= [HW] Parallel port keyboard adapter mode of operation,
3426 0 for XT, 1 for AT (default is AT).
3429 parport= [HW,PPT] Specify parallel ports. 0 disables.
3430 Format: { 0 | auto | 0xBBB[,IRQ[,DMA]] }
3431 Use 'auto' to force the driver to use any
3432 IRQ/DMA settings detected (the default is to
3433 ignore detected IRQ/DMA settings because of
3434 possible conflicts). You can specify the base
3435 address, IRQ, and DMA settings; IRQ and DMA
3436 should be numbers, or 'auto' (for using detected
3437 settings on that particular port), or 'nofifo'
3438 (to avoid using a FIFO even if it is detected).
3439 Parallel ports are assigned in the order they
3440 are specified on the command line, starting
3443 parport_init_mode= [HW,PPT]
3444 Configure VIA parallel port to operate in
3445 a specific mode. This is necessary on Pegasos
3446 computer where firmware has no options for setting
3447 up parallel port mode and sets it to spp.
3448 Currently this function knows 686a and 8231 chips.
3449 Format: [spp|ps2|epp|ecp|ecpepp]
3452 Halt all CPUs after the first oops has been printed for
3453 the specified number of seconds. This is to be used if
3454 your oopses keep scrolling off the screen.
3459 See header of drivers/block/paride/pcd.c.
3460 See also Documentation/admin-guide/blockdev/paride.rst.
3462 pci=option[,option...] [PCI] various PCI subsystem options.
3464 Some options herein operate on a specific device
3465 or a set of devices (<pci_dev>). These are
3466 specified in one of the following formats:
3468 [<domain>:]<bus>:<dev>.<func>[/<dev>.<func>]*
3469 pci:<vendor>:<device>[:<subvendor>:<subdevice>]
3471 Note: the first format specifies a PCI
3472 bus/device/function address which may change
3473 if new hardware is inserted, if motherboard
3474 firmware changes, or due to changes caused
3475 by other kernel parameters. If the
3476 domain is left unspecified, it is
3477 taken to be zero. Optionally, a path
3478 to a device through multiple device/function
3479 addresses can be specified after the base
3480 address (this is more robust against
3481 renumbering issues). The second format
3482 selects devices using IDs from the
3483 configuration space which may match multiple
3484 devices in the system.
3486 earlydump dump PCI config space before the kernel
3488 off [X86] don't probe for the PCI bus
3489 bios [X86-32] force use of PCI BIOS, don't access
3490 the hardware directly. Use this if your machine
3491 has a non-standard PCI host bridge.
3492 nobios [X86-32] disallow use of PCI BIOS, only direct
3493 hardware access methods are allowed. Use this
3494 if you experience crashes upon bootup and you
3495 suspect they are caused by the BIOS.
3496 conf1 [X86] Force use of PCI Configuration Access
3497 Mechanism 1 (config address in IO port 0xCF8,
3498 data in IO port 0xCFC, both 32-bit).
3499 conf2 [X86] Force use of PCI Configuration Access
3500 Mechanism 2 (IO port 0xCF8 is an 8-bit port for
3501 the function, IO port 0xCFA, also 8-bit, sets
3502 bus number. The config space is then accessed
3503 through ports 0xC000-0xCFFF).
3504 See http://wiki.osdev.org/PCI for more info
3505 on the configuration access mechanisms.
3506 noaer [PCIE] If the PCIEAER kernel config parameter is
3507 enabled, this kernel boot option can be used to
3508 disable the use of PCIE advanced error reporting.
3509 nodomains [PCI] Disable support for multiple PCI
3510 root domains (aka PCI segments, in ACPI-speak).
3511 nommconf [X86] Disable use of MMCONFIG for PCI
3513 check_enable_amd_mmconf [X86] check for and enable
3514 properly configured MMIO access to PCI
3515 config space on AMD family 10h CPU
3516 nomsi [MSI] If the PCI_MSI kernel config parameter is
3517 enabled, this kernel boot option can be used to
3518 disable the use of MSI interrupts system-wide.
3519 noioapicquirk [APIC] Disable all boot interrupt quirks.
3520 Safety option to keep boot IRQs enabled. This
3521 should never be necessary.
3522 ioapicreroute [APIC] Enable rerouting of boot IRQs to the
3523 primary IO-APIC for bridges that cannot disable
3524 boot IRQs. This fixes a source of spurious IRQs
3525 when the system masks IRQs.
3526 noioapicreroute [APIC] Disable workaround that uses the
3527 boot IRQ equivalent of an IRQ that connects to
3528 a chipset where boot IRQs cannot be disabled.
3529 The opposite of ioapicreroute.
3530 biosirq [X86-32] Use PCI BIOS calls to get the interrupt
3531 routing table. These calls are known to be buggy
3532 on several machines and they hang the machine
3533 when used, but on other computers it's the only
3534 way to get the interrupt routing table. Try
3535 this option if the kernel is unable to allocate
3536 IRQs or discover secondary PCI buses on your
3538 rom [X86] Assign address space to expansion ROMs.
3539 Use with caution as certain devices share
3540 address decoders between ROMs and other
3542 norom [X86] Do not assign address space to
3543 expansion ROMs that do not already have
3544 BIOS assigned address ranges.
3545 nobar [X86] Do not assign address space to the
3546 BARs that weren't assigned by the BIOS.
3547 irqmask=0xMMMM [X86] Set a bit mask of IRQs allowed to be
3548 assigned automatically to PCI devices. You can
3549 make the kernel exclude IRQs of your ISA cards
3551 pirqaddr=0xAAAAA [X86] Specify the physical address
3552 of the PIRQ table (normally generated
3553 by the BIOS) if it is outside the
3554 F0000h-100000h range.
3555 lastbus=N [X86] Scan all buses thru bus #N. Can be
3556 useful if the kernel is unable to find your
3557 secondary buses and you want to tell it
3558 explicitly which ones they are.
3559 assign-busses [X86] Always assign all PCI bus
3560 numbers ourselves, overriding
3561 whatever the firmware may have done.
3562 usepirqmask [X86] Honor the possible IRQ mask stored
3563 in the BIOS $PIR table. This is needed on
3564 some systems with broken BIOSes, notably
3565 some HP Pavilion N5400 and Omnibook XE3
3566 notebooks. This will have no effect if ACPI
3567 IRQ routing is enabled.
3568 noacpi [X86] Do not use ACPI for IRQ routing
3569 or for PCI scanning.
3570 use_crs [X86] Use PCI host bridge window information
3571 from ACPI. On BIOSes from 2008 or later, this
3572 is enabled by default. If you need to use this,
3573 please report a bug.
3574 nocrs [X86] Ignore PCI host bridge windows from ACPI.
3575 If you need to use this, please report a bug.
3576 routeirq Do IRQ routing for all PCI devices.
3577 This is normally done in pci_enable_device(),
3578 so this option is a temporary workaround
3579 for broken drivers that don't call it.
3580 skip_isa_align [X86] do not align io start addr, so can
3581 handle more pci cards
3582 noearly [X86] Don't do any early type 1 scanning.
3583 This might help on some broken boards which
3584 machine check when some devices' config space
3585 is read. But various workarounds are disabled
3586 and some IOMMU drivers will not work.
3587 bfsort Sort PCI devices into breadth-first order.
3588 This sorting is done to get a device
3589 order compatible with older (<= 2.4) kernels.
3590 nobfsort Don't sort PCI devices into breadth-first order.
3591 pcie_bus_tune_off Disable PCIe MPS (Max Payload Size)
3592 tuning and use the BIOS-configured MPS defaults.
3593 pcie_bus_safe Set every device's MPS to the largest value
3594 supported by all devices below the root complex.
3595 pcie_bus_perf Set device MPS to the largest allowable MPS
3596 based on its parent bus. Also set MRRS (Max
3597 Read Request Size) to the largest supported
3598 value (no larger than the MPS that the device
3599 or bus can support) for best performance.
3600 pcie_bus_peer2peer Set every device's MPS to 128B, which
3601 every device is guaranteed to support. This
3602 configuration allows peer-to-peer DMA between
3603 any pair of devices, possibly at the cost of
3604 reduced performance. This also guarantees
3605 that hot-added devices will work.
3606 cbiosize=nn[KMG] The fixed amount of bus space which is
3607 reserved for the CardBus bridge's IO window.
3608 The default value is 256 bytes.
3609 cbmemsize=nn[KMG] The fixed amount of bus space which is
3610 reserved for the CardBus bridge's memory
3611 window. The default value is 64 megabytes.
3614 [<order of align>@]<pci_dev>[; ...]
3615 Specifies alignment and device to reassign
3616 aligned memory resources. How to
3617 specify the device is described above.
3618 If <order of align> is not specified,
3619 PAGE_SIZE is used as alignment.
3620 A PCI-PCI bridge can be specified if resource
3621 windows need to be expanded.
3622 To specify the alignment for several
3623 instances of a device, the PCI vendor,
3624 device, subvendor, and subdevice may be
3625 specified, e.g., 12@pci:8086:9c22:103c:198f
3626 for 4096-byte alignment.
3627 ecrc= Enable/disable PCIe ECRC (transaction layer
3628 end-to-end CRC checking).
3629 bios: Use BIOS/firmware settings. This is the
3633 hpiosize=nn[KMG] The fixed amount of bus space which is
3634 reserved for hotplug bridge's IO window.
3635 Default size is 256 bytes.
3636 hpmmiosize=nn[KMG] The fixed amount of bus space which is
3637 reserved for hotplug bridge's MMIO window.
3638 Default size is 2 megabytes.
3639 hpmmioprefsize=nn[KMG] The fixed amount of bus space which is
3640 reserved for hotplug bridge's MMIO_PREF window.
3641 Default size is 2 megabytes.
3642 hpmemsize=nn[KMG] The fixed amount of bus space which is
3643 reserved for hotplug bridge's MMIO and
3645 Default size is 2 megabytes.
3646 hpbussize=nn The minimum amount of additional bus numbers
3647 reserved for buses below a hotplug bridge.
3649 realloc= Enable/disable reallocating PCI bridge resources
3650 if allocations done by BIOS are too small to
3651 accommodate resources required by all child
3653 off: Turn realloc off
3655 realloc same as realloc=on
3656 noari do not use PCIe ARI.
3657 noats [PCIE, Intel-IOMMU, AMD-IOMMU]
3658 do not use PCIe ATS (and IOMMU device IOTLB).
3659 pcie_scan_all Scan all possible PCIe devices. Otherwise we
3660 only look for one device below a PCIe downstream
3662 big_root_window Try to add a big 64bit memory window to the PCIe
3663 root complex on AMD CPUs. Some GFX hardware
3664 can resize a BAR to allow access to all VRAM.
3665 Adding the window is slightly risky (it may
3666 conflict with unreported devices), so this
3668 disable_acs_redir=<pci_dev>[; ...]
3669 Specify one or more PCI devices (in the format
3670 specified above) separated by semicolons.
3671 Each device specified will have the PCI ACS
3672 redirect capabilities forced off which will
3673 allow P2P traffic between devices through
3674 bridges without forcing it upstream. Note:
3675 this removes isolation between devices and
3676 may put more devices in an IOMMU group.
3677 force_floating [S390] Force usage of floating interrupts.
3678 nomio [S390] Do not use MIO instructions.
3680 pcie_aspm= [PCIE] Forcibly enable or disable PCIe Active State Power
3683 force Enable ASPM even on devices that claim not to support it.
3684 WARNING: Forcing ASPM on may cause system lockups.
3686 pcie_ports= [PCIE] PCIe port services handling:
3687 native Use native PCIe services (PME, AER, DPC, PCIe hotplug)
3688 even if the platform doesn't give the OS permission to
3689 use them. This may cause conflicts if the platform
3690 also tries to use these services.
3691 dpc-native Use native PCIe service for DPC only. May
3692 cause conflicts if firmware uses AER or DPC.
3693 compat Disable native PCIe services (PME, AER, DPC, PCIe
3696 pcie_port_pm= [PCIE] PCIe port power management handling:
3697 off Disable power management of all PCIe ports
3698 force Forcibly enable power management of all PCIe ports
3700 pcie_pme= [PCIE,PM] Native PCIe PME signaling options:
3701 nomsi Do not use MSI for native PCIe PME signaling (this makes
3702 all PCIe root ports use INTx for all services).
3704 pcmv= [HW,PCMCIA] BadgePAD 4
3708 Keep all power-domains already enabled by bootloader on,
3709 even if no driver has claimed them. This is useful
3710 for debug and development, but should not be
3711 needed on a platform with proper driver support.
3714 See Documentation/admin-guide/blockdev/paride.rst.
3716 pdcchassis= [PARISC,HW] Disable/Enable PDC Chassis Status codes at
3719 See arch/parisc/kernel/pdc_chassis.c
3721 percpu_alloc= Select which percpu first chunk allocator to use.
3722 Currently supported values are "embed" and "page".
3723 Archs may support subset or none of the selections.
3724 See comments in mm/percpu.c for details on each
3725 allocator. This parameter is primarily for debugging
3726 and performance comparison.
3729 See Documentation/admin-guide/blockdev/paride.rst.
3732 See Documentation/admin-guide/blockdev/paride.rst.
3734 pirq= [SMP,APIC] Manual mp-table setup
3735 See Documentation/x86/i386/IO-APIC.rst.
3737 plip= [PPT,NET] Parallel port network link
3738 Format: { parport<nr> | timid | 0 }
3739 See also Documentation/admin-guide/parport.rst.
3741 pmtmr= [X86] Manual setup of pmtmr I/O Port.
3742 Override pmtimer IOPort with a hex value.
3745 pm_debug_messages [SUSPEND,KNL]
3746 Enable suspend/resume debug messages during boot up.
3749 Enable PNP debug messages (depends on the
3750 CONFIG_PNP_DEBUG_MESSAGES option). Change at run-time
3751 via /sys/module/pnp/parameters/debug. We always show
3752 current resource usage; turning this on also shows
3753 possible settings and some assignment information.
3759 { on | off | curr | res | no-curr | no-res }
3762 [ISAPNP] Exclude IRQs for the autoconfiguration
3765 [ISAPNP] Exclude DMAs for the autoconfiguration
3767 pnp_reserve_io= [ISAPNP] Exclude I/O ports for the autoconfiguration
3768 Ranges are in pairs (I/O port base and size).
3771 [ISAPNP] Exclude memory regions for the
3773 Ranges are in pairs (memory base and size).
3775 ports= [IP_VS_FTP] IPVS ftp helper module
3777 Up to 8 (IP_VS_APP_MAX_PORTS) ports
3779 Format: <port>,<port>....
3781 powersave=off [PPC] This option disables power saving features.
3782 It specifically disables cpuidle and sets the
3783 platform machine description specific power_save
3784 function to NULL. On Idle the CPU just reduces
3787 ppc_strict_facility_enable
3788 [PPC] This option catches any kernel floating point,
3789 Altivec, VSX and SPE outside of regions specifically
3790 allowed (eg kernel_enable_fpu()/kernel_disable_fpu()).
3791 There is some performance impact when enabling this.
3795 Disable Hardware Transactional Memory
3797 print-fatal-signals=
3798 [KNL] debug: print fatal signals
3800 If enabled, warn about various signal handling
3801 related application anomalies: too many signals,
3802 too many POSIX.1 timers, fatal signals causing a
3805 If you hit the warning due to signal overflow,
3806 you might want to try "ulimit -i unlimited".
3810 printk.always_kmsg_dump=
3811 Trigger kmsg_dump for cases other than kernel oops or
3813 Format: <bool> (1/Y/y=enable, 0/N/n=disable)
3816 printk.devkmsg={on,off,ratelimit}
3817 Control writing to /dev/kmsg.
3818 on - unlimited logging to /dev/kmsg from userspace
3819 off - logging to /dev/kmsg disabled
3820 ratelimit - ratelimit the logging
3823 printk.time= Show timing data prefixed to each printk message line
3824 Format: <bool> (1/Y/y=enable, 0/N/n=disable)
3826 processor.max_cstate= [HW,ACPI]
3827 Limit processor to maximum C-state
3828 max_cstate=9 overrides any DMI blacklist limit.
3830 processor.nocst [HW,ACPI]
3831 Ignore the _CST method to determine C-states,
3832 instead using the legacy FADT method
3834 profile= [KNL] Enable kernel profiling via /proc/profile
3835 Format: [<profiletype>,]<number>
3836 Param: <profiletype>: "schedule", "sleep", or "kvm"
3837 [defaults to kernel profiling]
3838 Param: "schedule" - profile schedule points.
3839 Param: "sleep" - profile D-state sleeping (millisecs).
3840 Requires CONFIG_SCHEDSTATS
3841 Param: "kvm" - profile VM exits.
3842 Param: <number> - step/bucket size as a power of 2 for
3843 statistical time based profiling.
3845 prompt_ramdisk= [RAM] List of RAM disks to prompt for floppy disk
3847 See Documentation/admin-guide/blockdev/ramdisk.rst.
3849 prot_virt= [S390] enable hosting protected virtual machines
3850 isolated from the hypervisor (if hardware supports
3854 psi= [KNL] Enable or disable pressure stall information
3858 psmouse.proto= [HW,MOUSE] Highest PS2 mouse protocol extension to
3859 probe for; one of (bare|imps|exps|lifebook|any).
3860 psmouse.rate= [HW,MOUSE] Set desired mouse report rate, in reports
3862 psmouse.resetafter= [HW,MOUSE]
3863 Try to reset the device after so many bad packets
3866 [HW,MOUSE] Set desired mouse resolution, in dpi.
3867 psmouse.smartscroll=
3868 [HW,MOUSE] Controls Logitech smartscroll autorepeat.
3869 0 = disabled, 1 = enabled (default).
3871 pstore.backend= Specify the name of the pstore backend to use
3874 See Documentation/admin-guide/blockdev/paride.rst.
3876 pti= [X86_64] Control Page Table Isolation of user and
3877 kernel address spaces. Disabling this feature
3878 removes hardening, but improves performance of
3879 system calls and interrupts.
3881 on - unconditionally enable
3882 off - unconditionally disable
3883 auto - kernel detects whether your CPU model is
3884 vulnerable to issues that PTI mitigates
3886 Not specifying this option is equivalent to pti=auto.
3889 Equivalent to pti=off
3892 [KNL] Number of legacy pty's. Overwrites compiled-in
3895 quiet [KNL] Disable most log messages
3900 See Documentation/admin-guide/md.rst.
3902 ramdisk_size= [RAM] Sizes of RAM disks in kilobytes
3903 See Documentation/admin-guide/blockdev/ramdisk.rst.
3905 random.trust_cpu={on,off}
3906 [KNL] Enable or disable trusting the use of the
3907 CPU's random number generator (if available) to
3908 fully seed the kernel's CRNG. Default is controlled
3909 by CONFIG_RANDOM_TRUST_CPU.
3911 ras=option[,option,...] [KNL] RAS-specific options
3914 Disable the Correctable Errors Collector,
3915 see CONFIG_RAS_CEC help text.
3918 The argument is a cpu list, as described above,
3919 except that the string "all" can be used to
3920 specify every CPU on the system.
3922 In kernels built with CONFIG_RCU_NOCB_CPU=y, set
3923 the specified list of CPUs to be no-callback CPUs.
3924 Invocation of these CPUs' RCU callbacks will be
3925 offloaded to "rcuox/N" kthreads created for that
3926 purpose, where "x" is "p" for RCU-preempt, and
3927 "s" for RCU-sched, and "N" is the CPU number.
3928 This reduces OS jitter on the offloaded CPUs,
3929 which can be useful for HPC and real-time
3930 workloads. It can also improve energy efficiency
3931 for asymmetric multiprocessors.
3934 Rather than requiring that offloaded CPUs
3935 (specified by rcu_nocbs= above) explicitly
3936 awaken the corresponding "rcuoN" kthreads,
3937 make these kthreads poll for callbacks.
3938 This improves the real-time response for the
3939 offloaded CPUs by relieving them of the need to
3940 wake up the corresponding kthread, but degrades
3941 energy efficiency by requiring that the kthreads
3942 periodically wake up to do the polling.
3944 rcutree.blimit= [KNL]
3945 Set maximum number of finished RCU callbacks to
3946 process in one batch.
3948 rcutree.dump_tree= [KNL]
3949 Dump the structure of the rcu_node combining tree
3950 out at early boot. This is used for diagnostic
3951 purposes, to verify correct tree setup.
3953 rcutree.gp_cleanup_delay= [KNL]
3954 Set the number of jiffies to delay each step of
3955 RCU grace-period cleanup.
3957 rcutree.gp_init_delay= [KNL]
3958 Set the number of jiffies to delay each step of
3959 RCU grace-period initialization.
3961 rcutree.gp_preinit_delay= [KNL]
3962 Set the number of jiffies to delay each step of
3963 RCU grace-period pre-initialization, that is,
3964 the propagation of recent CPU-hotplug changes up
3965 the rcu_node combining tree.
3967 rcutree.use_softirq= [KNL]
3968 If set to zero, move all RCU_SOFTIRQ processing to
3969 per-CPU rcuc kthreads. Defaults to a non-zero
3970 value, meaning that RCU_SOFTIRQ is used by default.
3971 Specify rcutree.use_softirq=0 to use rcuc kthreads.
3973 rcutree.rcu_fanout_exact= [KNL]
3974 Disable autobalancing of the rcu_node combining
3975 tree. This is used by rcutorture, and might
3976 possibly be useful for architectures having high
3977 cache-to-cache transfer latencies.
3979 rcutree.rcu_fanout_leaf= [KNL]
3980 Change the number of CPUs assigned to each
3981 leaf rcu_node structure. Useful for very
3982 large systems, which will choose the value 64,
3983 and for NUMA systems with large remote-access
3984 latencies, which will choose a value aligned
3985 with the appropriate hardware boundaries.
3987 rcutree.jiffies_till_first_fqs= [KNL]
3988 Set delay from grace-period initialization to
3989 first attempt to force quiescent states.
3990 Units are jiffies, minimum value is zero,
3991 and maximum value is HZ.
3993 rcutree.jiffies_till_next_fqs= [KNL]
3994 Set delay between subsequent attempts to force
3995 quiescent states. Units are jiffies, minimum
3996 value is one, and maximum value is HZ.
3998 rcutree.jiffies_till_sched_qs= [KNL]
3999 Set required age in jiffies for a
4000 given grace period before RCU starts
4001 soliciting quiescent-state help from
4002 rcu_note_context_switch() and cond_resched().
4003 If not specified, the kernel will calculate
4004 a value based on the most recent settings
4005 of rcutree.jiffies_till_first_fqs
4006 and rcutree.jiffies_till_next_fqs.
4007 This calculated value may be viewed in
4008 rcutree.jiffies_to_sched_qs. Any attempt to set
4009 rcutree.jiffies_to_sched_qs will be cheerfully
4012 rcutree.kthread_prio= [KNL,BOOT]
4013 Set the SCHED_FIFO priority of the RCU per-CPU
4014 kthreads (rcuc/N). This value is also used for
4015 the priority of the RCU boost threads (rcub/N)
4016 and for the RCU grace-period kthreads (rcu_bh,
4017 rcu_preempt, and rcu_sched). If RCU_BOOST is
4018 set, valid values are 1-99 and the default is 1
4019 (the least-favored priority). Otherwise, when
4020 RCU_BOOST is not set, valid values are 0-99 and
4021 the default is zero (non-realtime operation).
4023 rcutree.rcu_nocb_gp_stride= [KNL]
4024 Set the number of NOCB callback kthreads in
4025 each group, which defaults to the square root
4026 of the number of CPUs. Larger numbers reduce
4027 the wakeup overhead on the global grace-period
4028 kthread, but increases that same overhead on
4029 each group's NOCB grace-period kthread.
4031 rcutree.qhimark= [KNL]
4032 Set threshold of queued RCU callbacks beyond which
4033 batch limiting is disabled.
4035 rcutree.qlowmark= [KNL]
4036 Set threshold of queued RCU callbacks below which
4037 batch limiting is re-enabled.
4039 rcutree.qovld= [KNL]
4040 Set threshold of queued RCU callbacks beyond which
4041 RCU's force-quiescent-state scan will aggressively
4042 enlist help from cond_resched() and sched IPIs to
4043 help CPUs more quickly reach quiescent states.
4044 Set to less than zero to make this be set based
4045 on rcutree.qhimark at boot time and to zero to
4046 disable more aggressive help enlistment.
4048 rcutree.rcu_idle_gp_delay= [KNL]
4049 Set wakeup interval for idle CPUs that have
4050 RCU callbacks (RCU_FAST_NO_HZ=y).
4052 rcutree.rcu_idle_lazy_gp_delay= [KNL]
4053 Set wakeup interval for idle CPUs that have
4054 only "lazy" RCU callbacks (RCU_FAST_NO_HZ=y).
4055 Lazy RCU callbacks are those which RCU can
4056 prove do nothing more than free memory.
4058 rcutree.rcu_kick_kthreads= [KNL]
4059 Cause the grace-period kthread to get an extra
4060 wake_up() if it sleeps three times longer than
4061 it should at force-quiescent-state time.
4062 This wake_up() will be accompanied by a
4063 WARN_ONCE() splat and an ftrace_dump().
4065 rcutree.sysrq_rcu= [KNL]
4066 Commandeer a sysrq key to dump out Tree RCU's
4067 rcu_node tree with an eye towards determining
4068 why a new grace period has not yet started.
4070 rcuperf.gp_async= [KNL]
4071 Measure performance of asynchronous
4072 grace-period primitives such as call_rcu().
4074 rcuperf.gp_async_max= [KNL]
4075 Specify the maximum number of outstanding
4076 callbacks per writer thread. When a writer
4077 thread exceeds this limit, it invokes the
4078 corresponding flavor of rcu_barrier() to allow
4079 previously posted callbacks to drain.
4081 rcuperf.gp_exp= [KNL]
4082 Measure performance of expedited synchronous
4083 grace-period primitives.
4085 rcuperf.holdoff= [KNL]
4086 Set test-start holdoff period. The purpose of
4087 this parameter is to delay the start of the
4088 test until boot completes in order to avoid
4091 rcuperf.kfree_rcu_test= [KNL]
4092 Set to measure performance of kfree_rcu() flooding.
4094 rcuperf.kfree_nthreads= [KNL]
4095 The number of threads running loops of kfree_rcu().
4097 rcuperf.kfree_alloc_num= [KNL]
4098 Number of allocations and frees done in an iteration.
4100 rcuperf.kfree_loops= [KNL]
4101 Number of loops doing rcuperf.kfree_alloc_num number
4102 of allocations and frees.
4104 rcuperf.nreaders= [KNL]
4105 Set number of RCU readers. The value -1 selects
4106 N, where N is the number of CPUs. A value
4107 "n" less than -1 selects N-n+1, where N is again
4108 the number of CPUs. For example, -2 selects N
4109 (the number of CPUs), -3 selects N+1, and so on.
4110 A value of "n" less than or equal to -N selects
4113 rcuperf.nwriters= [KNL]
4114 Set number of RCU writers. The values operate
4115 the same as for rcuperf.nreaders.
4116 N, where N is the number of CPUs
4118 rcuperf.perf_type= [KNL]
4119 Specify the RCU implementation to test.
4121 rcuperf.shutdown= [KNL]
4122 Shut the system down after performance tests
4123 complete. This is useful for hands-off automated
4126 rcuperf.verbose= [KNL]
4127 Enable additional printk() statements.
4129 rcuperf.writer_holdoff= [KNL]
4130 Write-side holdoff between grace periods,
4131 in microseconds. The default of zero says
4134 rcutorture.fqs_duration= [KNL]
4135 Set duration of force_quiescent_state bursts
4138 rcutorture.fqs_holdoff= [KNL]
4139 Set holdoff time within force_quiescent_state bursts
4142 rcutorture.fqs_stutter= [KNL]
4143 Set wait time between force_quiescent_state bursts
4146 rcutorture.fwd_progress= [KNL]
4147 Enable RCU grace-period forward-progress testing
4148 for the types of RCU supporting this notion.
4150 rcutorture.fwd_progress_div= [KNL]
4151 Specify the fraction of a CPU-stall-warning
4152 period to do tight-loop forward-progress testing.
4154 rcutorture.fwd_progress_holdoff= [KNL]
4155 Number of seconds to wait between successive
4156 forward-progress tests.
4158 rcutorture.fwd_progress_need_resched= [KNL]
4159 Enclose cond_resched() calls within checks for
4160 need_resched() during tight-loop forward-progress
4163 rcutorture.gp_cond= [KNL]
4164 Use conditional/asynchronous update-side
4165 primitives, if available.
4167 rcutorture.gp_exp= [KNL]
4168 Use expedited update-side primitives, if available.
4170 rcutorture.gp_normal= [KNL]
4171 Use normal (non-expedited) asynchronous
4172 update-side primitives, if available.
4174 rcutorture.gp_sync= [KNL]
4175 Use normal (non-expedited) synchronous
4176 update-side primitives, if available. If all
4177 of rcutorture.gp_cond=, rcutorture.gp_exp=,
4178 rcutorture.gp_normal=, and rcutorture.gp_sync=
4179 are zero, rcutorture acts as if is interpreted
4180 they are all non-zero.
4182 rcutorture.n_barrier_cbs= [KNL]
4183 Set callbacks/threads for rcu_barrier() testing.
4185 rcutorture.nfakewriters= [KNL]
4186 Set number of concurrent RCU writers. These just
4187 stress RCU, they don't participate in the actual
4188 test, hence the "fake".
4190 rcutorture.nreaders= [KNL]
4191 Set number of RCU readers. The value -1 selects
4192 N-1, where N is the number of CPUs. A value
4193 "n" less than -1 selects N-n-2, where N is again
4194 the number of CPUs. For example, -2 selects N
4195 (the number of CPUs), -3 selects N+1, and so on.
4197 rcutorture.object_debug= [KNL]
4198 Enable debug-object double-call_rcu() testing.
4200 rcutorture.onoff_holdoff= [KNL]
4201 Set time (s) after boot for CPU-hotplug testing.
4203 rcutorture.onoff_interval= [KNL]
4204 Set time (jiffies) between CPU-hotplug operations,
4205 or zero to disable CPU-hotplug testing.
4207 rcutorture.shuffle_interval= [KNL]
4208 Set task-shuffle interval (s). Shuffling tasks
4209 allows some CPUs to go into dyntick-idle mode
4210 during the rcutorture test.
4212 rcutorture.shutdown_secs= [KNL]
4213 Set time (s) after boot system shutdown. This
4214 is useful for hands-off automated testing.
4216 rcutorture.stall_cpu= [KNL]
4217 Duration of CPU stall (s) to test RCU CPU stall
4218 warnings, zero to disable.
4220 rcutorture.stall_cpu_block= [KNL]
4221 Sleep while stalling if set. This will result
4222 in warnings from preemptible RCU in addition
4223 to any other stall-related activity.
4225 rcutorture.stall_cpu_holdoff= [KNL]
4226 Time to wait (s) after boot before inducing stall.
4228 rcutorture.stall_cpu_irqsoff= [KNL]
4229 Disable interrupts while stalling if set.
4231 rcutorture.stall_gp_kthread= [KNL]
4232 Duration (s) of forced sleep within RCU
4233 grace-period kthread to test RCU CPU stall
4234 warnings, zero to disable. If both stall_cpu
4235 and stall_gp_kthread are specified, the
4236 kthread is starved first, then the CPU.
4238 rcutorture.stat_interval= [KNL]
4239 Time (s) between statistics printk()s.
4241 rcutorture.stutter= [KNL]
4242 Time (s) to stutter testing, for example, specifying
4243 five seconds causes the test to run for five seconds,
4244 wait for five seconds, and so on. This tests RCU's
4245 ability to transition abruptly to and from idle.
4247 rcutorture.test_boost= [KNL]
4248 Test RCU priority boosting? 0=no, 1=maybe, 2=yes.
4249 "Maybe" means test if the RCU implementation
4250 under test support RCU priority boosting.
4252 rcutorture.test_boost_duration= [KNL]
4253 Duration (s) of each individual boost test.
4255 rcutorture.test_boost_interval= [KNL]
4256 Interval (s) between each boost test.
4258 rcutorture.test_no_idle_hz= [KNL]
4259 Test RCU's dyntick-idle handling. See also the
4260 rcutorture.shuffle_interval parameter.
4262 rcutorture.torture_type= [KNL]
4263 Specify the RCU implementation to test.
4265 rcutorture.verbose= [KNL]
4266 Enable additional printk() statements.
4268 rcupdate.rcu_cpu_stall_ftrace_dump= [KNL]
4269 Dump ftrace buffer after reporting RCU CPU
4272 rcupdate.rcu_cpu_stall_suppress= [KNL]
4273 Suppress RCU CPU stall warning messages.
4275 rcupdate.rcu_cpu_stall_suppress_at_boot= [KNL]
4276 Suppress RCU CPU stall warning messages and
4277 rcutorture writer stall warnings that occur
4278 during early boot, that is, during the time
4279 before the init task is spawned.
4281 rcupdate.rcu_cpu_stall_timeout= [KNL]
4282 Set timeout for RCU CPU stall warning messages.
4284 rcupdate.rcu_expedited= [KNL]
4285 Use expedited grace-period primitives, for
4286 example, synchronize_rcu_expedited() instead
4287 of synchronize_rcu(). This reduces latency,
4288 but can increase CPU utilization, degrade
4289 real-time latency, and degrade energy efficiency.
4290 No effect on CONFIG_TINY_RCU kernels.
4292 rcupdate.rcu_normal= [KNL]
4293 Use only normal grace-period primitives,
4294 for example, synchronize_rcu() instead of
4295 synchronize_rcu_expedited(). This improves
4296 real-time latency, CPU utilization, and
4297 energy efficiency, but can expose users to
4298 increased grace-period latency. This parameter
4299 overrides rcupdate.rcu_expedited. No effect on
4300 CONFIG_TINY_RCU kernels.
4302 rcupdate.rcu_normal_after_boot= [KNL]
4303 Once boot has completed (that is, after
4304 rcu_end_inkernel_boot() has been invoked), use
4305 only normal grace-period primitives. No effect
4306 on CONFIG_TINY_RCU kernels.
4308 rcupdate.rcu_task_ipi_delay= [KNL]
4309 Set time in jiffies during which RCU tasks will
4310 avoid sending IPIs, starting with the beginning
4311 of a given grace period. Setting a large
4312 number avoids disturbing real-time workloads,
4313 but lengthens grace periods.
4315 rcupdate.rcu_task_stall_timeout= [KNL]
4316 Set timeout in jiffies for RCU task stall warning
4317 messages. Disable with a value less than or equal
4320 rcupdate.rcu_self_test= [KNL]
4321 Run the RCU early boot self tests
4325 Run specified binary instead of /init from the ramdisk,
4326 used for early userspace startup. See initrd.
4329 force - Override the decision by the kernel to hide the
4330 advertisement of RDRAND support (this affects
4331 certain AMD processors because of buggy BIOS
4332 support, specifically around the suspend/resume
4336 Turn on/off individual RDT features. List is:
4337 cmt, mbmtotal, mbmlocal, l3cat, l3cdp, l2cat, l2cdp,
4339 E.g. to turn on cmt and turn off mba use:
4343 Format (x86 or x86_64):
4344 [w[arm] | c[old] | h[ard] | s[oft] | g[pio]] \
4346 [[,]b[ios] | a[cpi] | k[bd] | t[riple] | e[fi] | p[ci]] \
4348 Where reboot_mode is one of warm (soft) or cold (hard) or gpio
4349 (prefix with 'panic_' to set mode for panic
4351 reboot_type is one of bios, acpi, kbd, triple, efi, or pci,
4352 reboot_force is either force or not specified,
4353 reboot_cpu is s[mp]#### with #### being the processor
4354 to be used for rebooting.
4357 [KNL, SMP] Set scheduler's default relax_domain_level.
4358 See Documentation/admin-guide/cgroup-v1/cpusets.rst.
4360 reserve= [KNL,BUGS] Force kernel to ignore I/O ports or memory
4361 Format: <base1>,<size1>[,<base2>,<size2>,...]
4362 Reserve I/O ports or memory so the kernel won't use
4363 them. If <base> is less than 0x10000, the region
4364 is assumed to be I/O ports; otherwise it is memory.
4366 reservetop= [X86-32]
4368 Reserves a hole at the top of the kernel virtual
4373 Set the amount of memory to reserve for BIOS at
4374 the bottom of the address space.
4376 reset_devices [KNL] Force drivers to reset the underlying device
4377 during initialization.
4380 Specify the partition device for software suspend
4382 {/dev/<dev> | PARTUUID=<uuid> | <int>:<int> | <hex>}
4384 resume_offset= [SWSUSP]
4385 Specify the offset from the beginning of the partition
4386 given by "resume=" at which the swap header is located,
4387 in <PAGE_SIZE> units (needed only for swap files).
4388 See Documentation/power/swsusp-and-swap-files.rst
4390 resumedelay= [HIBERNATION] Delay (in seconds) to pause before attempting to
4391 read the resume files
4393 resumewait [HIBERNATION] Wait (indefinitely) for resume device to show up.
4394 Useful for devices that are detected asynchronously
4395 (e.g. USB and MMC devices).
4397 hibernate= [HIBERNATION]
4398 noresume Don't check if there's a hibernation image
4399 present during boot.
4400 nocompress Don't compress/decompress hibernation images.
4401 no Disable hibernation and resume.
4402 protect_image Turn on image protection during restoration
4403 (that will set all pages holding image data
4404 during restoration read-only).
4406 retain_initrd [RAM] Keep initrd memory after extraction
4408 rfkill.default_state=
4409 0 "airplane mode". All wifi, bluetooth, wimax, gps, fm,
4410 etc. communication is blocked by default.
4413 rfkill.master_switch_mode=
4414 0 The "airplane mode" button does nothing.
4415 1 The "airplane mode" button toggles between everything
4416 blocked and the previous configuration.
4417 2 The "airplane mode" button toggles between everything
4418 blocked and everything unblocked.
4420 rhash_entries= [KNL,NET]
4421 Set number of hash buckets for route cache
4424 [KNL] Disable ring 3 MONITOR/MWAIT feature on supported
4427 ro [KNL] Mount root device read-only on boot
4430 on Mark read-only kernel memory as read-only (default).
4431 off Leave read-only kernel memory writable for debugging.
4434 Enable the uart passthrough on the designated usb port
4435 on Rockchip SoCs. When active, the signals of the
4436 debug-uart get routed to the D+ and D- pins of the usb
4437 port and the regular usb controller gets disabled.
4439 root= [KNL] Root filesystem
4440 See name_to_dev_t comment in init/do_mounts.c.
4442 rootdelay= [KNL] Delay (in seconds) to pause before attempting to
4443 mount the root filesystem
4445 rootflags= [KNL] Set root filesystem mount option string
4447 rootfstype= [KNL] Set root filesystem type
4449 rootwait [KNL] Wait (indefinitely) for root device to show up.
4450 Useful for devices that are detected asynchronously
4451 (e.g. USB and MMC devices).
4453 rproc_mem=nn[KMG][@address]
4454 [KNL,ARM,CMA] Remoteproc physical memory block.
4455 Memory area to be used by remote processor image,
4458 rw [KNL] Mount root device read-write on boot
4460 S [KNL] Run init in single mode
4462 s390_iommu= [HW,S390]
4463 Set s390 IOTLB flushing mode
4465 With strict flushing every unmap operation will result in
4466 an IOTLB flush. Default is lazy flushing before reuse,
4470 See drivers/net/irda/sa1100_ir.c.
4472 sbni= [NET] Granch SBNI12 leased line adapter
4474 sched_debug [KNL] Enables verbose scheduler debug messages.
4476 schedstats= [KNL,X86] Enable or disable scheduled statistics.
4477 Allowed values are enable and disable. This feature
4478 incurs a small amount of overhead in the scheduler
4479 but is useful for debugging and performance tuning.
4481 sched_thermal_decay_shift=
4482 [KNL, SMP] Set a decay shift for scheduler thermal
4483 pressure signal. Thermal pressure signal follows the
4484 default decay period of other scheduler pelt
4485 signals(usually 32 ms but configurable). Setting
4486 sched_thermal_decay_shift will left shift the decay
4487 period for the thermal pressure signal by the shift
4489 i.e. with the default pelt decay period of 32 ms
4490 sched_thermal_decay_shift thermal pressure decay pr
4494 Format: integer between 0 and 10
4497 skew_tick= [KNL] Offset the periodic timer tick per cpu to mitigate
4498 xtime_lock contention on larger systems, and/or RCU lock
4499 contention on all systems with CONFIG_MAXSMP set.
4500 Format: { "0" | "1" }
4501 0 -- disable. (may be 1 via CONFIG_CMDLINE="skew_tick=1"
4503 Note: increases power consumption, thus should only be
4504 enabled if running jitter sensitive (HPC/RT) workloads.
4506 security= [SECURITY] Choose a legacy "major" security module to
4507 enable at boot. This has been deprecated by the
4510 selinux= [SELINUX] Disable or enable SELinux at boot time.
4511 Format: { "0" | "1" }
4512 See security/selinux/Kconfig help text.
4517 apparmor= [APPARMOR] Disable or enable AppArmor at boot time
4518 Format: { "0" | "1" }
4519 See security/apparmor/Kconfig help text
4522 Default value is set via kernel config option.
4524 serialnumber [BUGS=X86-32]
4527 Maximal number of shapers.
4535 Disable merging of slabs with similar size. May be
4536 necessary if there is some reason to distinguish
4537 allocs to different slabs, especially in hardened
4538 environments where the risk of heap overflows and
4539 layout control by attackers can usually be
4540 frustrated by disabling merging. This will reduce
4541 most of the exposure of a heap attack to a single
4542 cache (risks via metadata attacks are mostly
4543 unchanged). Debug options disable merging on their
4545 For more information see Documentation/vm/slub.rst.
4547 slab_max_order= [MM, SLAB]
4548 Determines the maximum allowed order for slabs.
4549 A high setting may cause OOMs due to memory
4550 fragmentation. Defaults to 1 for systems with
4551 more than 32MB of RAM, 0 otherwise.
4553 slub_debug[=options[,slabs]] [MM, SLUB]
4554 Enabling slub_debug allows one to determine the
4555 culprit if slab objects become corrupted. Enabling
4556 slub_debug can create guard zones around objects and
4557 may poison objects when not in use. Also tracks the
4558 last alloc / free. For more information see
4559 Documentation/vm/slub.rst.
4561 slub_memcg_sysfs= [MM, SLUB]
4562 Determines whether to enable sysfs directories for
4563 memory cgroup sub-caches. 1 to enable, 0 to disable.
4564 The default is determined by CONFIG_SLUB_MEMCG_SYSFS_ON.
4565 Enabling this can lead to a very high number of debug
4566 directories and files being created under
4569 slub_max_order= [MM, SLUB]
4570 Determines the maximum allowed order for slabs.
4571 A high setting may cause OOMs due to memory
4572 fragmentation. For more information see
4573 Documentation/vm/slub.rst.
4575 slub_min_objects= [MM, SLUB]
4576 The minimum number of objects per slab. SLUB will
4577 increase the slab order up to slub_max_order to
4578 generate a sufficiently large slab able to contain
4579 the number of objects indicated. The higher the number
4580 of objects the smaller the overhead of tracking slabs
4581 and the less frequently locks need to be acquired.
4582 For more information see Documentation/vm/slub.rst.
4584 slub_min_order= [MM, SLUB]
4585 Determines the minimum page order for slabs. Must be
4586 lower than slub_max_order.
4587 For more information see Documentation/vm/slub.rst.
4589 slub_nomerge [MM, SLUB]
4590 Same with slab_nomerge. This is supported for legacy.
4591 See slab_nomerge for more information.
4594 Format: <io1>[,<io2>[,...,<io8>]]
4596 smsc-ircc2.nopnp [HW] Don't use PNP to discover SMC devices
4597 smsc-ircc2.ircc_cfg= [HW] Device configuration I/O port
4598 smsc-ircc2.ircc_sir= [HW] SIR base I/O port
4599 smsc-ircc2.ircc_fir= [HW] FIR base I/O port
4600 smsc-ircc2.ircc_irq= [HW] IRQ line
4601 smsc-ircc2.ircc_dma= [HW] DMA channel
4602 smsc-ircc2.ircc_transceiver= [HW] Transceiver type:
4603 0: Toshiba Satellite 1800 (GP data pin select)
4604 1: Fast pin select (default)
4607 smt [KNL,S390] Set the maximum number of threads (logical
4608 CPUs) to use per physical CPU on systems capable of
4609 symmetric multithreading (SMT). Will be capped to the
4610 actual hardware limit.
4612 Default: -1 (no limit)
4615 [KNL] Should the soft-lockup detector generate panics.
4618 A nonzero value instructs the soft-lockup detector
4619 to panic the machine when a soft-lockup occurs. It is
4620 also controlled by the kernel.softlockup_panic sysctl
4621 and CONFIG_BOOTPARAM_SOFTLOCKUP_PANIC, which is the
4622 respective build-time switch to that functionality.
4624 softlockup_all_cpu_backtrace=
4625 [KNL] Should the soft-lockup detector generate
4626 backtraces on all cpus.
4629 sonypi.*= [HW] Sony Programmable I/O Control Device driver
4630 See Documentation/admin-guide/laptops/sonypi.rst
4632 spectre_v2= [X86] Control mitigation of Spectre variant 2
4633 (indirect branch speculation) vulnerability.
4634 The default operation protects the kernel from
4637 on - unconditionally enable, implies
4639 off - unconditionally disable, implies
4641 auto - kernel detects whether your CPU model is
4644 Selecting 'on' will, and 'auto' may, choose a
4645 mitigation method at run time according to the
4646 CPU, the available microcode, the setting of the
4647 CONFIG_RETPOLINE configuration option, and the
4648 compiler with which the kernel was built.
4650 Selecting 'on' will also enable the mitigation
4651 against user space to user space task attacks.
4653 Selecting 'off' will disable both the kernel and
4654 the user space protections.
4656 Specific mitigations can also be selected manually:
4658 retpoline - replace indirect branches
4659 retpoline,generic - google's original retpoline
4660 retpoline,amd - AMD-specific minimal thunk
4662 Not specifying this option is equivalent to
4666 [X86] Control mitigation of Spectre variant 2
4667 (indirect branch speculation) vulnerability between
4670 on - Unconditionally enable mitigations. Is
4671 enforced by spectre_v2=on
4673 off - Unconditionally disable mitigations. Is
4674 enforced by spectre_v2=off
4676 prctl - Indirect branch speculation is enabled,
4677 but mitigation can be enabled via prctl
4678 per thread. The mitigation control state
4679 is inherited on fork.
4682 - Like "prctl" above, but only STIBP is
4683 controlled per thread. IBPB is issued
4684 always when switching between different user
4688 - Same as "prctl" above, but all seccomp
4689 threads will enable the mitigation unless
4690 they explicitly opt out.
4693 - Like "seccomp" above, but only STIBP is
4694 controlled per thread. IBPB is issued
4695 always when switching between different
4696 user space processes.
4698 auto - Kernel selects the mitigation depending on
4699 the available CPU features and vulnerability.
4702 If CONFIG_SECCOMP=y then "seccomp", otherwise "prctl"
4704 Not specifying this option is equivalent to
4705 spectre_v2_user=auto.
4707 spec_store_bypass_disable=
4708 [HW] Control Speculative Store Bypass (SSB) Disable mitigation
4709 (Speculative Store Bypass vulnerability)
4711 Certain CPUs are vulnerable to an exploit against a
4712 a common industry wide performance optimization known
4713 as "Speculative Store Bypass" in which recent stores
4714 to the same memory location may not be observed by
4715 later loads during speculative execution. The idea
4716 is that such stores are unlikely and that they can
4717 be detected prior to instruction retirement at the
4718 end of a particular speculation execution window.
4720 In vulnerable processors, the speculatively forwarded
4721 store can be used in a cache side channel attack, for
4722 example to read memory to which the attacker does not
4723 directly have access (e.g. inside sandboxed code).
4725 This parameter controls whether the Speculative Store
4726 Bypass optimization is used.
4728 On x86 the options are:
4730 on - Unconditionally disable Speculative Store Bypass
4731 off - Unconditionally enable Speculative Store Bypass
4732 auto - Kernel detects whether the CPU model contains an
4733 implementation of Speculative Store Bypass and
4734 picks the most appropriate mitigation. If the
4735 CPU is not vulnerable, "off" is selected. If the
4736 CPU is vulnerable the default mitigation is
4737 architecture and Kconfig dependent. See below.
4738 prctl - Control Speculative Store Bypass per thread
4739 via prctl. Speculative Store Bypass is enabled
4740 for a process by default. The state of the control
4741 is inherited on fork.
4742 seccomp - Same as "prctl" above, but all seccomp threads
4743 will disable SSB unless they explicitly opt out.
4745 Default mitigations:
4746 X86: If CONFIG_SECCOMP=y "seccomp", otherwise "prctl"
4748 On powerpc the options are:
4750 on,auto - On Power8 and Power9 insert a store-forwarding
4751 barrier on kernel entry and exit. On Power7
4752 perform a software flush on kernel entry and
4756 Not specifying this option is equivalent to
4757 spec_store_bypass_disable=auto.
4759 spia_io_base= [HW,MTD]
4765 [X86] Enable split lock detection
4767 When enabled (and if hardware support is present), atomic
4768 instructions that access data across cache line
4769 boundaries will result in an alignment check exception.
4773 warn - the kernel will emit rate limited warnings
4774 about applications triggering the #AC
4775 exception. This mode is the default on CPUs
4776 that supports split lock detection.
4778 fatal - the kernel will send SIGBUS to applications
4779 that trigger the #AC exception.
4781 If an #AC exception is hit in the kernel or in
4782 firmware (i.e. not while executing in user mode)
4783 the kernel will oops in either "warn" or "fatal"
4786 srcutree.counter_wrap_check [KNL]
4787 Specifies how frequently to check for
4788 grace-period sequence counter wrap for the
4789 srcu_data structure's ->srcu_gp_seq_needed field.
4790 The greater the number of bits set in this kernel
4791 parameter, the less frequently counter wrap will
4792 be checked for. Note that the bottom two bits
4795 srcutree.exp_holdoff [KNL]
4796 Specifies how many nanoseconds must elapse
4797 since the end of the last SRCU grace period for
4798 a given srcu_struct until the next normal SRCU
4799 grace period will be considered for automatic
4800 expediting. Set to zero to disable automatic
4804 Speculative Store Bypass Disable control
4806 On CPUs that are vulnerable to the Speculative
4807 Store Bypass vulnerability and offer a
4808 firmware based mitigation, this parameter
4809 indicates how the mitigation should be used:
4811 force-on: Unconditionally enable mitigation for
4812 for both kernel and userspace
4813 force-off: Unconditionally disable mitigation for
4814 for both kernel and userspace
4815 kernel: Always enable mitigation in the
4816 kernel, and offer a prctl interface
4817 to allow userspace to register its
4818 interest in being mitigated too.
4820 stack_guard_gap= [MM]
4821 override the default stack gap protection. The value
4822 is in page units and it defines how many pages prior
4823 to (for stacks growing down) resp. after (for stacks
4824 growing up) the main stack are reserved for no other
4825 mapping. Default value is 256 pages.
4828 Enabled the stack tracer on boot up.
4830 stacktrace_filter=[function-list]
4831 [FTRACE] Limit the functions that the stack tracer
4832 will trace at boot up. function-list is a comma separated
4833 list of functions. This list can be changed at run
4834 time by the stack_trace_filter file in the debugfs
4835 tracing directory. Note, this enables stack tracing
4836 and the stacktrace above is not needed.
4840 Set the STI (builtin display/keyboard on the HP-PARISC
4841 machines) console (graphic card) which should be used
4842 as the initial boot-console.
4843 See also comment in drivers/video/console/sticore.c.
4846 See comment in drivers/video/console/sticore.c.
4849 Format: bpp:<bpp1>[:<bpp2>[:<bpp3>...]]
4851 sunrpc.min_resvport=
4852 sunrpc.max_resvport=
4854 SunRPC servers often require that client requests
4855 originate from a privileged port (i.e. a port in the
4856 range 0 < portnr < 1024).
4857 An administrator who wishes to reserve some of these
4858 ports for other uses may adjust the range that the
4859 kernel's sunrpc client considers to be privileged
4860 using these two parameters to set the minimum and
4861 maximum port values.
4863 sunrpc.svc_rpc_per_connection_limit=
4865 Limit the number of requests that the server will
4866 process in parallel from a single connection.
4867 The default value is 0 (no limit).
4871 Control how the NFS server code allocates CPUs to
4872 service thread pools. Depending on how many NICs
4873 you have and where their interrupts are bound, this
4874 option will affect which CPUs will do NFS serving.
4875 Note: this parameter cannot be changed while the
4876 NFS server is running.
4878 auto the server chooses an appropriate mode
4879 automatically using heuristics
4880 global a single global pool contains all CPUs
4881 percpu one pool for each CPU
4882 pernode one pool for each NUMA node (equivalent
4883 to global on non-NUMA machines)
4885 sunrpc.tcp_slot_table_entries=
4886 sunrpc.udp_slot_table_entries=
4888 Sets the upper limit on the number of simultaneous
4889 RPC calls that can be sent from the client to a
4890 server. Increasing these values may allow you to
4891 improve throughput, but will also increase the
4892 amount of memory reserved for use by the client.
4894 suspend.pm_test_delay=
4896 Sets the number of seconds to remain in a suspend test
4897 mode before resuming the system (see
4898 /sys/power/pm_test). Only available when CONFIG_PM_DEBUG
4899 is set. Default value is 5.
4902 Format: { on | off | y | n | 1 | 0 }
4903 This parameter controls use of the Protected
4904 Execution Facility on pSeries.
4907 [KNL] Enable accounting of swap in memory resource
4908 controller if no parameter or 1 is given or disable
4909 it if 0 is given (See Documentation/admin-guide/cgroup-v1/memory.rst)
4911 swiotlb= [ARM,IA-64,PPC,MIPS,X86]
4912 Format: { <int> | force | noforce }
4913 <int> -- Number of I/O TLB slabs
4914 force -- force using of bounce buffers even if they
4915 wouldn't be automatically used by the kernel
4916 noforce -- Never use bounce buffers (for debugging)
4920 sysfs.deprecated=0|1 [KNL]
4921 Enable/disable old style sysfs layout for old udev
4922 on older distributions. When this option is enabled
4923 very new udev will not work anymore. When this option
4924 is disabled (or CONFIG_SYSFS_DEPRECATED not compiled)
4925 in older udev will not work anymore.
4926 Default depends on CONFIG_SYSFS_DEPRECATED_V2 set in
4927 the kernel configuration.
4929 sysrq_always_enabled
4931 Ignore sysrq setting - this boot parameter will
4932 neutralize any effect of /proc/sys/kernel/sysrq.
4933 Useful for debugging.
4935 tcpmhash_entries= [KNL,NET]
4936 Set the number of tcp_metrics_hash slots.
4937 Default value is 8192 or 16384 depending on total
4938 ram pages. This is used to specify the TCP metrics
4939 cache size. See Documentation/networking/ip-sysctl.txt
4940 "tcp_no_metrics_save" section for more details.
4944 test_suspend= [SUSPEND][,N]
4945 Specify "mem" (for Suspend-to-RAM) or "standby" (for
4946 standby suspend) or "freeze" (for suspend type freeze)
4947 as the system sleep state during system startup with
4948 the optional capability to repeat N number of times.
4949 The system is woken from this state using a
4950 wakeup-capable RTC alarm.
4952 thash_entries= [KNL,NET]
4953 Set number of hash buckets for TCP connection
4955 thermal.act= [HW,ACPI]
4956 -1: disable all active trip points in all thermal zones
4957 <degrees C>: override all lowest active trip points
4959 thermal.crt= [HW,ACPI]
4960 -1: disable all critical trip points in all thermal zones
4961 <degrees C>: override all critical trip points
4963 thermal.nocrt= [HW,ACPI]
4964 Set to disable actions on ACPI thermal zone
4965 critical and hot trip points.
4967 thermal.off= [HW,ACPI]
4968 1: disable ACPI thermal control
4970 thermal.psv= [HW,ACPI]
4971 -1: disable all passive trip points
4972 <degrees C>: override all passive trip points to this
4975 thermal.tzp= [HW,ACPI]
4976 Specify global default ACPI thermal zone polling rate
4977 <deci-seconds>: poll all this frequency
4978 0: no polling (default)
4981 Force threading of all interrupt handlers except those
4982 marked explicitly IRQF_NO_THREAD.
4986 Specify if the kernel should make use of the cpu
4987 topology information if the hardware supports this.
4988 The scheduler will make use of this information and
4989 e.g. base its process migration decisions on it.
4992 topology_updates= [KNL, PPC, NUMA]
4994 Specify if the kernel should ignore (off)
4995 topology updates sent by the hypervisor to this
4998 torture.disable_onoff_at_boot= [KNL]
4999 Prevent the CPU-hotplug component of torturing
5000 until after init has spawned.
5004 tpm_suspend_pcr=[HW,TPM]
5005 Format: integer pcr id
5006 Specify that at suspend time, the tpm driver
5007 should extend the specified pcr with zeros,
5008 as a workaround for some chips which fail to
5009 flush the last written pcr on TPM_SaveState.
5010 This will guarantee that all the other pcrs
5013 trace_buf_size=nn[KMG]
5014 [FTRACE] will set tracing buffer size on each cpu.
5016 trace_event=[event-list]
5017 [FTRACE] Set and start specified trace events in order
5018 to facilitate early boot debugging. The event-list is a
5019 comma separated list of trace events to enable. See
5020 also Documentation/trace/events.rst
5022 trace_options=[option-list]
5023 [FTRACE] Enable or disable tracer options at boot.
5024 The option-list is a comma delimited list of options
5025 that can be enabled or disabled just as if you were
5026 to echo the option name into
5028 /sys/kernel/debug/tracing/trace_options
5030 For example, to enable stacktrace option (to dump the
5031 stack trace of each event), add to the command line:
5033 trace_options=stacktrace
5035 See also Documentation/trace/ftrace.rst "trace options"
5039 Have the tracepoints sent to printk as well as the
5040 tracing ring buffer. This is useful for early boot up
5041 where the system hangs or reboots and does not give the
5042 option for reading the tracing buffer or performing a
5043 ftrace_dump_on_oops.
5045 To turn off having tracepoints sent to printk,
5046 echo 0 > /proc/sys/kernel/tracepoint_printk
5047 Note, echoing 1 into this file without the
5048 tracepoint_printk kernel cmdline option has no effect.
5052 Having tracepoints sent to printk() and activating high
5053 frequency tracepoints such as irq or sched, can cause
5054 the system to live lock.
5057 [FTRACE] enable this option to disable tracing when a
5058 warning is hit. This turns off "tracing_on". Tracing can
5059 be enabled again by echoing '1' into the "tracing_on"
5060 file located in /sys/kernel/debug/tracing/
5062 This option is useful, as it disables the trace before
5063 the WARNING dump is called, which prevents the trace to
5064 be filled with content caused by the warning output.
5066 This option can also be set at run time via the sysctl
5067 option: kernel/traceoff_on_warning
5069 transparent_hugepage=
5071 Format: [always|madvise|never]
5072 Can be used to control the default behavior of the system
5073 with respect to transparent hugepages.
5074 See Documentation/admin-guide/mm/transhuge.rst
5077 tsc= Disable clocksource stability checks for TSC.
5079 [x86] reliable: mark tsc clocksource as reliable, this
5080 disables clocksource verification at runtime, as well
5081 as the stability checks done at bootup. Used to enable
5082 high-resolution timer mode on older hardware, and in
5083 virtualized environment.
5084 [x86] noirqtime: Do not use TSC to do irq accounting.
5085 Used to run time disable IRQ_TIME_ACCOUNTING on any
5086 platforms where RDTSC is slow and this accounting
5088 [x86] unstable: mark the TSC clocksource as unstable, this
5089 marks the TSC unconditionally unstable at bootup and
5090 avoids any further wobbles once the TSC watchdog notices.
5091 [x86] nowatchdog: disable clocksource watchdog. Used
5092 in situations with strict latency requirements (where
5093 interruptions from clocksource watchdog are not
5096 tsx= [X86] Control Transactional Synchronization
5097 Extensions (TSX) feature in Intel processors that
5098 support TSX control.
5100 This parameter controls the TSX feature. The options are:
5102 on - Enable TSX on the system. Although there are
5103 mitigations for all known security vulnerabilities,
5104 TSX has been known to be an accelerator for
5105 several previous speculation-related CVEs, and
5106 so there may be unknown security risks associated
5107 with leaving it enabled.
5109 off - Disable TSX on the system. (Note that this
5110 option takes effect only on newer CPUs which are
5111 not vulnerable to MDS, i.e., have
5112 MSR_IA32_ARCH_CAPABILITIES.MDS_NO=1 and which get
5113 the new IA32_TSX_CTRL MSR through a microcode
5114 update. This new MSR allows for the reliable
5115 deactivation of the TSX functionality.)
5117 auto - Disable TSX if X86_BUG_TAA is present,
5118 otherwise enable TSX on the system.
5120 Not specifying this option is equivalent to tsx=off.
5122 See Documentation/admin-guide/hw-vuln/tsx_async_abort.rst
5125 tsx_async_abort= [X86,INTEL] Control mitigation for the TSX Async
5126 Abort (TAA) vulnerability.
5128 Similar to Micro-architectural Data Sampling (MDS)
5129 certain CPUs that support Transactional
5130 Synchronization Extensions (TSX) are vulnerable to an
5131 exploit against CPU internal buffers which can forward
5132 information to a disclosure gadget under certain
5135 In vulnerable processors, the speculatively forwarded
5136 data can be used in a cache side channel attack, to
5137 access data to which the attacker does not have direct
5140 This parameter controls the TAA mitigation. The
5143 full - Enable TAA mitigation on vulnerable CPUs
5146 full,nosmt - Enable TAA mitigation and disable SMT on
5147 vulnerable CPUs. If TSX is disabled, SMT
5148 is not disabled because CPU is not
5149 vulnerable to cross-thread TAA attacks.
5150 off - Unconditionally disable TAA mitigation
5152 On MDS-affected machines, tsx_async_abort=off can be
5153 prevented by an active MDS mitigation as both vulnerabilities
5154 are mitigated with the same mechanism so in order to disable
5155 this mitigation, you need to specify mds=off too.
5157 Not specifying this option is equivalent to
5158 tsx_async_abort=full. On CPUs which are MDS affected
5159 and deploy MDS mitigation, TAA mitigation is not
5160 required and doesn't provide any additional
5164 Documentation/admin-guide/hw-vuln/tsx_async_abort.rst
5166 turbografx.map[2|3]= [HW,JOY]
5167 TurboGraFX parallel port interface
5169 <port#>,<js1>,<js2>,<js3>,<js4>,<js5>,<js6>,<js7>
5170 See also Documentation/input/devices/joystick-parport.rst
5172 udbg-immortal [PPC] When debugging early kernel crashes that
5173 happen after console_init() and before a proper
5174 console driver takes over, this boot options might
5175 help "seeing" what's going on.
5177 uhash_entries= [KNL,NET]
5178 Set number of hash buckets for UDP/UDP-Lite connections
5181 [USB] Ignore overcurrent events (default N).
5182 Some badly-designed motherboards generate lots of
5183 bogus events, for ports that aren't wired to
5184 anything. Set this parameter to avoid log spamming.
5185 Note that genuine overcurrent events won't be
5189 [X86] Cause panic on unknown NMI.
5191 usbcore.authorized_default=
5192 [USB] Default USB device authorization:
5193 (default -1 = authorized except for wireless USB,
5194 0 = not authorized, 1 = authorized, 2 = authorized
5195 if device connected to internal port)
5197 usbcore.autosuspend=
5198 [USB] The autosuspend time delay (in seconds) used
5199 for newly-detected USB devices (default 2). This
5200 is the time required before an idle device will be
5201 autosuspended. Devices for which the delay is set
5202 to a negative value won't be autosuspended at all.
5204 usbcore.usbfs_snoop=
5205 [USB] Set to log all usbfs traffic (default 0 = off).
5207 usbcore.usbfs_snoop_max=
5208 [USB] Maximum number of bytes to snoop in each URB
5211 usbcore.blinkenlights=
5212 [USB] Set to cycle leds on hubs (default 0 = off).
5214 usbcore.old_scheme_first=
5215 [USB] Start with the old device initialization
5216 scheme (default 0 = off).
5218 usbcore.usbfs_memory_mb=
5219 [USB] Memory limit (in MB) for buffers allocated by
5220 usbfs (default = 16, 0 = max = 2047).
5222 usbcore.use_both_schemes=
5223 [USB] Try the other device initialization scheme
5224 if the first one fails (default 1 = enabled).
5226 usbcore.initial_descriptor_timeout=
5227 [USB] Specifies timeout for the initial 64-byte
5228 USB_REQ_GET_DESCRIPTOR request in milliseconds
5229 (default 5000 = 5.0 seconds).
5231 usbcore.nousb [USB] Disable the USB subsystem
5234 [USB] A list of quirk entries to augment the built-in
5235 usb core quirk list. List entries are separated by
5236 commas. Each entry has the form
5237 VendorID:ProductID:Flags. The IDs are 4-digit hex
5238 numbers and Flags is a set of letters. Each letter
5239 will change the built-in quirk; setting it if it is
5240 clear and clearing it if it is set. The letters have
5241 the following meanings:
5242 a = USB_QUIRK_STRING_FETCH_255 (string
5243 descriptors must not be fetched using
5245 b = USB_QUIRK_RESET_RESUME (device can't resume
5246 correctly so reset it instead);
5247 c = USB_QUIRK_NO_SET_INTF (device can't handle
5248 Set-Interface requests);
5249 d = USB_QUIRK_CONFIG_INTF_STRINGS (device can't
5250 handle its Configuration or Interface
5252 e = USB_QUIRK_RESET (device can't be reset
5253 (e.g morph devices), don't use reset);
5254 f = USB_QUIRK_HONOR_BNUMINTERFACES (device has
5255 more interface descriptions than the
5256 bNumInterfaces count, and can't handle
5257 talking to these interfaces);
5258 g = USB_QUIRK_DELAY_INIT (device needs a pause
5259 during initialization, after we read
5260 the device descriptor);
5261 h = USB_QUIRK_LINEAR_UFRAME_INTR_BINTERVAL (For
5262 high speed and super speed interrupt
5263 endpoints, the USB 2.0 and USB 3.0 spec
5264 require the interval in microframes (1
5265 microframe = 125 microseconds) to be
5266 calculated as interval = 2 ^
5268 Devices with this quirk report their
5269 bInterval as the result of this
5270 calculation instead of the exponent
5271 variable used in the calculation);
5272 i = USB_QUIRK_DEVICE_QUALIFIER (device can't
5273 handle device_qualifier descriptor
5275 j = USB_QUIRK_IGNORE_REMOTE_WAKEUP (device
5276 generates spurious wakeup, ignore
5277 remote wakeup capability);
5278 k = USB_QUIRK_NO_LPM (device can't handle Link
5280 l = USB_QUIRK_LINEAR_FRAME_INTR_BINTERVAL
5281 (Device reports its bInterval as linear
5282 frames instead of the USB 2.0
5284 m = USB_QUIRK_DISCONNECT_SUSPEND (Device needs
5285 to be disconnected before suspend to
5286 prevent spurious wakeup);
5287 n = USB_QUIRK_DELAY_CTRL_MSG (Device needs a
5288 pause after every control message);
5289 o = USB_QUIRK_HUB_SLOW_RESET (Hub needs extra
5290 delay after resetting its port);
5291 Example: quirks=0781:5580:bk,0a5c:5834:gij
5294 [USBHID] The interval which mice are to be polled at.
5297 [USBHID] The interval which joysticks are to be polled at.
5300 [USBHID] The interval which keyboards are to be polled at.
5302 usb-storage.delay_use=
5303 [UMS] The delay in seconds before a new device is
5304 scanned for Logical Units (default 1).
5307 [UMS] A list of quirks entries to supplement or
5308 override the built-in unusual_devs list. List
5309 entries are separated by commas. Each entry has
5310 the form VID:PID:Flags where VID and PID are Vendor
5311 and Product ID values (4-digit hex numbers) and
5312 Flags is a set of characters, each corresponding
5313 to a common usb-storage quirk flag as follows:
5314 a = SANE_SENSE (collect more than 18 bytes
5315 of sense data, not on uas);
5316 b = BAD_SENSE (don't collect more than 18
5317 bytes of sense data, not on uas);
5318 c = FIX_CAPACITY (decrease the reported
5319 device capacity by one sector);
5320 d = NO_READ_DISC_INFO (don't use
5321 READ_DISC_INFO command, not on uas);
5322 e = NO_READ_CAPACITY_16 (don't use
5323 READ_CAPACITY_16 command);
5324 f = NO_REPORT_OPCODES (don't use report opcodes
5326 g = MAX_SECTORS_240 (don't transfer more than
5327 240 sectors at a time, uas only);
5328 h = CAPACITY_HEURISTICS (decrease the
5329 reported device capacity by one
5330 sector if the number is odd);
5331 i = IGNORE_DEVICE (don't bind to this
5333 j = NO_REPORT_LUNS (don't use report luns
5335 l = NOT_LOCKABLE (don't try to lock and
5336 unlock ejectable media, not on uas);
5337 m = MAX_SECTORS_64 (don't transfer more
5338 than 64 sectors = 32 KB at a time,
5340 n = INITIAL_READ10 (force a retry of the
5341 initial READ(10) command, not on uas);
5342 o = CAPACITY_OK (accept the capacity
5343 reported by the device, not on uas);
5344 p = WRITE_CACHE (the device cache is ON
5345 by default, not on uas);
5346 r = IGNORE_RESIDUE (the device reports
5347 bogus residue values, not on uas);
5348 s = SINGLE_LUN (the device has only one
5350 t = NO_ATA_1X (don't allow ATA(12) and ATA(16)
5351 commands, uas only);
5352 u = IGNORE_UAS (don't bind to the uas driver);
5353 w = NO_WP_DETECT (don't test whether the
5354 medium is write-protected).
5355 y = ALWAYS_SYNC (issue a SYNCHRONIZE_CACHE
5356 even if the device claims no cache,
5358 Example: quirks=0419:aaf5:rl,0421:0433:rc
5360 user_debug= [KNL,ARM]
5362 See arch/arm/Kconfig.debug help text.
5363 1 - undefined instruction events
5365 4 - invalid data aborts
5368 Example: user_debug=31
5371 [X86] Flags controlling user PTE allocations.
5373 nohigh = do not allocate PTE pages in
5374 HIGHMEM regardless of setting
5378 On X86_32, this is an alias for vdso32=. Otherwise:
5380 vdso=1: enable VDSO (the default)
5381 vdso=0: disable VDSO mapping
5383 vdso32= [X86] Control the 32-bit vDSO
5384 vdso32=1: enable 32-bit VDSO
5385 vdso32=0 or vdso32=2: disable 32-bit VDSO
5387 See the help text for CONFIG_COMPAT_VDSO for more
5388 details. If CONFIG_COMPAT_VDSO is set, the default is
5389 vdso32=0; otherwise, the default is vdso32=1.
5391 For compatibility with older kernels, vdso32=2 is an
5394 Try vdso32=0 if you encounter an error that says:
5395 dl_main: Assertion `(void *) ph->p_vaddr == _rtld_local._dl_sysinfo_dso' failed!
5398 vector=percpu: enable percpu vector domain
5400 video= [FB] Frame buffer configuration
5401 See Documentation/fb/modedb.rst.
5403 video.brightness_switch_enabled= [0,1]
5404 If set to 1, on receiving an ACPI notify event
5405 generated by hotkey, video driver will adjust brightness
5406 level and then send out the event to user space through
5407 the allocated input device; If set to 0, video driver
5408 will only send out the event without touching backlight
5413 [VMMIO] Memory mapped virtio (platform) device.
5415 <size>@<baseaddr>:<irq>[:<id>]
5417 <size> := size (can use standard suffixes
5419 <baseaddr> := physical base address
5420 <irq> := interrupt number (as passed to
5422 <id> := (optional) platform device id
5424 virtio_mmio.device=1K@0x100b0000:48:7
5426 Can be used multiple times for multiple devices.
5428 vga= [BOOT,X86-32] Select a particular video mode
5429 See Documentation/x86/boot.rst and
5430 Documentation/admin-guide/svga.rst.
5431 Use vga=ask for menu.
5432 This is actually a boot loader parameter; the value is
5433 passed to the kernel using a special protocol.
5435 vm_debug[=options] [KNL] Available with CONFIG_DEBUG_VM=y.
5436 May slow down system boot speed, especially when
5437 enabled on systems with a large amount of memory.
5438 All options are enabled by default, and this
5439 interface is meant to allow for selectively
5440 enabling or disabling specific virtual memory
5443 Available options are:
5444 P Enable page structure init time poisoning
5445 - Disable all of the above options
5447 vmalloc=nn[KMG] [KNL,BOOT] Forces the vmalloc area to have an exact
5448 size of <nn>. This can be used to increase the
5449 minimum size (128MB on x86). It can also be used to
5450 decrease the size and leave more room for directly
5453 vmcp_cma=nn[MG] [KNL,S390]
5454 Sets the memory size reserved for contiguous memory
5455 allocations for the vmcp device driver.
5457 vmhalt= [KNL,S390] Perform z/VM CP command after system halt.
5460 vmpanic= [KNL,S390] Perform z/VM CP command after kernel panic.
5463 vmpoff= [KNL,S390] Perform z/VM CP command after power off.
5467 Controls the behavior of vsyscalls (i.e. calls to
5468 fixed addresses of 0xffffffffff600x00 from legacy
5469 code). Most statically-linked binaries and older
5470 versions of glibc use these calls. Because these
5471 functions are at fixed addresses, they make nice
5472 targets for exploits that can control RIP.
5474 emulate [default] Vsyscalls turn into traps and are
5475 emulated reasonably safely. The vsyscall
5478 xonly Vsyscalls turn into traps and are
5479 emulated reasonably safely. The vsyscall
5480 page is not readable.
5482 none Vsyscalls don't work at all. This makes
5483 them quite hard to use for exploits but
5484 might break your system.
5486 vt.color= [VT] Default text color.
5487 Format: 0xYX, X = foreground, Y = background.
5488 Default: 0x07 = light gray on black.
5490 vt.cur_default= [VT] Default cursor shape.
5491 Format: 0xCCBBAA, where AA, BB, and CC are the same as
5492 the parameters of the <Esc>[?A;B;Cc escape sequence;
5493 see VGA-softcursor.txt. Default: 2 = underline.
5495 vt.default_blu= [VT]
5496 Format: <blue0>,<blue1>,<blue2>,...,<blue15>
5497 Change the default blue palette of the console.
5498 This is a 16-member array composed of values
5501 vt.default_grn= [VT]
5502 Format: <green0>,<green1>,<green2>,...,<green15>
5503 Change the default green palette of the console.
5504 This is a 16-member array composed of values
5507 vt.default_red= [VT]
5508 Format: <red0>,<red1>,<red2>,...,<red15>
5509 Change the default red palette of the console.
5510 This is a 16-member array composed of values
5516 Set system-wide default UTF-8 mode for all tty's.
5517 Default is 1, i.e. UTF-8 mode is enabled for all
5518 newly opened terminals.
5520 vt.global_cursor_default=
5523 Set system-wide default for whether a cursor
5524 is shown on new VTs. Default is -1,
5525 i.e. cursors will be created by default unless
5526 overridden by individual drivers. 0 will hide
5527 cursors, 1 will display them.
5529 vt.italic= [VT] Default color for italic text; 0-15.
5532 vt.underline= [VT] Default color for underlined text; 0-15.
5535 watchdog timers [HW,WDT] For information on watchdog timers,
5536 see Documentation/watchdog/watchdog-parameters.rst
5537 or other driver-specific files in the
5538 Documentation/watchdog/ directory.
5542 Set the hard lockup detector stall duration
5543 threshold in seconds. The soft lockup detector
5544 threshold is set to twice the value. A value of 0
5545 disables both lockup detectors. Default is 10
5548 workqueue.watchdog_thresh=
5549 If CONFIG_WQ_WATCHDOG is configured, workqueue can
5550 warn stall conditions and dump internal state to
5551 help debugging. 0 disables workqueue stall
5552 detection; otherwise, it's the stall threshold
5553 duration in seconds. The default value is 30 and
5554 it can be updated at runtime by writing to the
5555 corresponding sysfs file.
5557 workqueue.disable_numa
5558 By default, all work items queued to unbound
5559 workqueues are affine to the NUMA nodes they're
5560 issued on, which results in better behavior in
5561 general. If NUMA affinity needs to be disabled for
5562 whatever reason, this option can be used. Note
5563 that this also can be controlled per-workqueue for
5564 workqueues visible under /sys/bus/workqueue/.
5566 workqueue.power_efficient
5567 Per-cpu workqueues are generally preferred because
5568 they show better performance thanks to cache
5569 locality; unfortunately, per-cpu workqueues tend to
5570 be more power hungry than unbound workqueues.
5572 Enabling this makes the per-cpu workqueues which
5573 were observed to contribute significantly to power
5574 consumption unbound, leading to measurably lower
5575 power usage at the cost of small performance
5578 The default value of this parameter is determined by
5579 the config option CONFIG_WQ_POWER_EFFICIENT_DEFAULT.
5581 workqueue.debug_force_rr_cpu
5582 Workqueue used to implicitly guarantee that work
5583 items queued without explicit CPU specified are put
5584 on the local CPU. This guarantee is no longer true
5585 and while local CPU is still preferred work items
5586 may be put on foreign CPUs. This debug option
5587 forces round-robin CPU selection to flush out
5588 usages which depend on the now broken guarantee.
5589 When enabled, memory and cache locality will be
5592 x2apic_phys [X86-64,APIC] Use x2apic physical mode instead of
5593 default x2apic cluster mode on platforms
5596 x86_intel_mid_timer= [X86-32,APBT]
5597 Choose timer option for x86 Intel MID platform.
5598 Two valid options are apbt timer only and lapic timer
5599 plus one apbt timer for broadcast timer.
5600 x86_intel_mid_timer=apbt_only | lapic_and_apbt
5602 xen_512gb_limit [KNL,X86-64,XEN]
5603 Restricts the kernel running paravirtualized under Xen
5604 to use only up to 512 GB of RAM. The reason to do so is
5605 crash analysis tools and Xen tools for doing domain
5606 save/restore/migration must be enabled to handle larger
5609 xen_emul_unplug= [HW,X86,XEN]
5610 Unplug Xen emulated devices
5611 Format: [unplug0,][unplug1]
5612 ide-disks -- unplug primary master IDE devices
5613 aux-ide-disks -- unplug non-primary-master IDE devices
5614 nics -- unplug network devices
5615 all -- unplug all emulated devices (NICs and IDE disks)
5616 unnecessary -- unplugging emulated devices is
5617 unnecessary even if the host did not respond to
5619 never -- do not unplug even if version check succeeds
5621 xen_legacy_crash [X86,XEN]
5622 Crash from Xen panic notifier, without executing late
5623 panic() code such as dumping handler.
5625 xen_nopvspin [X86,XEN]
5626 Disables the ticketlock slowpath using Xen PV
5630 Disables the PV optimizations forcing the HVM guest to
5631 run as generic HVM guest with no PV drivers.
5632 This option is obsoleted by the "nopv" option, which
5633 has equivalent effect for XEN platform.
5635 xen_scrub_pages= [XEN]
5636 Boolean option to control scrubbing pages before giving them back
5637 to Xen, for use by other domains. Can be also changed at runtime
5638 with /sys/devices/system/xen_memory/xen_memory0/scrub_pages.
5639 Default value controlled with CONFIG_XEN_SCRUB_PAGES_DEFAULT.
5641 xen_timer_slop= [X86-64,XEN]
5642 Set the timer slop (in nanoseconds) for the virtual Xen
5643 timers (default is 100000). This adjusts the minimum
5644 delta of virtualized Xen timers, where lower values
5645 improve timer resolution at the expense of processing
5646 more timer interrupts.
5648 nopv= [X86,XEN,KVM,HYPER_V,VMWARE]
5649 Disables the PV optimizations forcing the guest to run
5650 as generic guest with no PV drivers. Currently support
5651 XEN HVM, KVM, HYPER_V and VMWARE guest.
5653 xirc2ps_cs= [NET,PCMCIA]
5655 <irq>,<irq_mask>,<io>,<full_duplex>,<do_sound>,<lockup_hack>[,<irq2>[,<irq3>[,<irq4>]]]
5658 By default on POWER9 and above, the kernel will
5659 natively use the XIVE interrupt controller. This option
5660 allows the fallback firmware mode to be used:
5662 off Fallback to firmware control of XIVE interrupt
5663 controller on both pseries and powernv
5664 platforms. Only useful on POWER9 and above.
5666 xhci-hcd.quirks [USB,KNL]
5667 A hex value specifying bitmask with supplemental xhci
5668 host controller quirks. Meaning of each bit can be
5669 consulted in header drivers/usb/host/xhci.h.
5672 Format: { early | on | rw | ro | off }
5673 Controls if xmon debugger is enabled. Default is off.
5674 Passing only "xmon" is equivalent to "xmon=early".
5675 early Call xmon as early as possible on boot; xmon
5676 debugger is called from setup_arch().
5677 on xmon debugger hooks will be installed so xmon
5678 is only called on a kernel crash. Default mode,
5679 i.e. either "ro" or "rw" mode, is controlled
5680 with CONFIG_XMON_DEFAULT_RO_MODE.
5681 rw xmon debugger hooks will be installed so xmon
5682 is called only on a kernel crash, mode is write,
5683 meaning SPR registers, memory and, other data
5684 can be written using xmon commands.
5685 ro same as "rw" option above but SPR registers,
5686 memory, and other data can't be written using
5688 off xmon is disabled.